DEPARTMENT OF CHEMICAL ENGINEERING AND BIOTECHNOLOGY
Microwave-Assisted Pyrolysis of HDPE
using an Activated Carbon Bed
Author:
Alan Russell
Clare College
Supervisor:
Professor Howard Chase
Magdalene College
September 2012
This dissertation is submitted for the degree of Doctor of Philosophy

Preface
This work was carried out in the Department of Chemical Engineering and Biotechno-
logy, University of Cambridge, from October 2008 to August 2012, and is submitted in
fulfilment of the requirements of the degree of Doctor of Philosophy. This dissertation
is the result of my own work and includes nothing that is the outcome of work done
in collaboration except where specifically indicated in the text. None of the work
contained in this dissertation, or any part thereof, has been submitted for any other
degree at the University of Cambridge or any other institution. This dissertation does
not exceed 65,000 words, and contains fewer than 150 figures.
Alan Russell
20th September 2012
i

Summary
Plastics play an enormous role in modern manufacturing, but the extraction and
refining of raw materials, followed by the synthesis of plastics themselves, represents
an enormous energy investment into a product that is all too often simply “thrown
away” into a landfill after a single use. Microwave-assisted pyrolysis is a recycling
technique that allows the recovery of chemical value from plastic waste by breaking
down polymers into useful smaller hydrocarbons using microwave heat in the absence
of oxygen. This dissertation examines the use of a catalytic activated carbon bed in this
procedure, using high density polyethylene (HDPE) as a model plastic.
Initial tests with the batch input of HDPE produced a condensed pyrolysis oil
comprising 35.5–45.3% aromatics, with the remainder primarily short-chain aliphatics.
This oil was approximately three times lighter than that produced in the absence of
catalyst, with a narrower range of molecular masses that matched those of the liquid
transport fuels petrol and diesel (C5–C21). The non-condensable gases that resulted
were short-chain aliphatics that could be used as feedstock for the creation of new
chemicals (such as virgin HDPE), or fuels such as natural gas and LPG.
The development of apparatus capable of adding sample in a continuous fashion
enabled the processing of larger quantities of HDPE, and resulted in condensed
products with a significantly higher aromatic content (>80% at 450°C), and which
encompassed a somewhat narrower range of molecular masses compared with those
produced in the batch mode; this was due to differences in kinetics and residence time
that resulted from the different modes of sample introduction. As a result of processing
larger quantities of HDPE it became apparent that the activated carbon deactivated
over time, with a bed able to process around 3.5 times its mass in HDPE at 450°C
before any significant changes in output products occurred.
The decomposition of HDPE proceeds via thermal scission and radical-mediated
mechanisms; high energy surface active sites facilitate the transfer of hydrogen and
radicals, and this enhances overall cracking and lowers the activation energy for
the formation of aromatics. Analysis of material deposited on the surface of the
activated carbon confirmed that deactivation occurred through coking, with both
cracking and deactivation thought to be enhanced by the formation of microwave-
induced microplasmas.
Overall, the microwave-assisted pyrolysis of HDPE using activated carbon produces
a much narrower range of more valuable products compared with non-catalytic
processing. While the process is not likely to be economic in its current form owing
to the relatively rapid deactivation of the activated carbon, future configurations
incorporating online reactivation may be able to economically provide a second
use cycle for these materials, avoiding expending energy to extract and process
increasingly scarce new raw material from the surface of the earth.
iii

Acknowledgements
I owe a debt of thanks to many people for the support and kindness they have
shown me throughout my time in Cambridge. I would like to thank my supervisor,
Professor Howard Chase, for his generous provision of tutelage, expertise, and support
throughout the production of this work. A further thank you to my colleagues and the
department support staff for their extremely skilled time and advice.
I would also like to acknowledge the generous financial support that I have received
from the Cambridge Commonwealth Trust, The Higher Education Funding Council for
England, Clare College, The Department of Chemical Engineering and Biotechnology,
and The Cambridge Philosophical Society, without which my time in Cambridge would
not have been possible.
The multitude of amazing people I have met over the last four years have truly made
my time at Cambridge special. Thank you to all my friends with whom I have shared the
good times and the bad, who have contributed time, thoughts, support, and probably a
little too much food, drink, fun, and cupcakes!
My family has always been an incredible source of unwavering encouragement in
my life, and antipodean distance has only made me realise and appreciate their love
and support all the more. I owe my mother, Anne, an extra special thank you, not least
for her willingness to employ her incredible eagle eye in hunting out orthographical
errors from this dissertation.
v

Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 The Problem and Opportunity of Plastic Waste . . . . . . . . . . . . . . . . . . 1
1.2 Current Plastics Recycling Techniques . . . . . . . . . . . . . . . . . . . . . . . 2
1.2.1 Mechanical Recycling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2.2 Energy Recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2.3 Feedstock Recycling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2.3.1 Pyrolysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2.3.2 Catalytic Pyrolysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3 Microwave-Assisted Pyrolysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3.1 Microwave Heating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3.1.1 Advantages of Microwave Heating . . . . . . . . . . . . . . . . . 5
1.3.1.2 Disadvantages of Microwave Heating . . . . . . . . . . . . . . . 6
1.3.2 Microwave-Assisted Pyrolysis of Plastics using a Carbon Bed . . . . . 6
1.3.2.1 Method for Microwave-Assisted Pyrolysis of Plastics Using a
Carbon Bed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.3.3 Carbon as a Reactor Bed Material . . . . . . . . . . . . . . . . . . . . . . 7
1.3.3.1 Coke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.3.3.2 Activated Carbon . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.4 Petrochemicals and Hydrocarbon Fuels . . . . . . . . . . . . . . . . . . . . . . 9
1.4.1 Cost Savings and Efficiency Benefits of Feedstock Recycling . . . . . . 10
1.5 High Density Polyethylene (HDPE) . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.6 Potential Benefits from the Microwave-Assisted Pyrolysis of HDPE . . . . . . 11
1.7 Motivation and Objectives for the Study . . . . . . . . . . . . . . . . . . . . . . 12
1.8 Outline of Thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2 Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.1 Carbon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.1.1 Allotropes of Carbon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.1.2 Non-Graphitisable Carbon . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.2 Porosity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.3 Activated Carbon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.3.1 Material Origin and Carbonisation . . . . . . . . . . . . . . . . . . . . . . 18
2.3.2 Activation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.3.2.1 Thermal Activation . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.3.2.2 Chemical Activation . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.3.3 Applications of Activated Carbon . . . . . . . . . . . . . . . . . . . . . . . 21
2.3.3.1 Commercial Applications of Activated Carbon . . . . . . . . . . 21
2.3.3.2 Activated Carbon as a Catalyst Support . . . . . . . . . . . . . . 21
2.3.4 Activated Carbon as a Catalyst . . . . . . . . . . . . . . . . . . . . . . . . 22
2.3.4.1 Catalytic Hydrogen Production using Activated Carbon . . . . 23
2.4 Microwave Heating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.4.1 Dielectric Properties of Materials and Influence on Microwave Re-
ceptivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.4.1.1 Carbonaceous Materials as Microwave Receptors . . . . . . . 25
2.4.1.2 Power and Penetration Depth . . . . . . . . . . . . . . . . . . . . 26
2.4.2 Microwave Heating Modes . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.4.2.1 Single-Mode Microwave Heating . . . . . . . . . . . . . . . . . . 26
2.4.2.2 Multi-Mode Microwave Heating . . . . . . . . . . . . . . . . . . 27
2.4.3 Microwave vs. Thermal Effects . . . . . . . . . . . . . . . . . . . . . . . . 27
2.4.3.1 Microplasmas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
vii
viii CONTENTS
2.5 Microwave-Assisted Pyrolysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.5.1 Processing of Problematic Waste . . . . . . . . . . . . . . . . . . . . . . 29
2.5.1.1 Scrap Tyres . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.5.1.2 Used Automotive Engine Oil . . . . . . . . . . . . . . . . . . . . . 29
2.5.1.3 Sewage Sludge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.5.2 Microwave-Assisted Pyrolysis of Biomass Waste . . . . . . . . . . . . . 30
2.5.3 Upgrading Naturally Occurring Hydrocarbon Sources . . . . . . . . . 30
2.5.4 Microwave-Assisted Pyrolysis with an Activated Carbon Bed . . . . . . 31
2.6 Pyrolysis of Plastics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.6.1 Pyrolysis of Polyethylene . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.6.1.1 Microwave-Assisted Pyrolysis of HDPE . . . . . . . . . . . . . . 34
2.6.1.2 HDPE Pyrolysis Overview . . . . . . . . . . . . . . . . . . . . . . 34
2.6.2 Pyrolysis of Other Plastics . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
2.6.2.1 Polypropylene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
2.6.2.2 Polystyrene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
2.6.2.3 Poly(vinyl chloride) . . . . . . . . . . . . . . . . . . . . . . . . . . 36
2.6.2.4 Poly(methyl methacrylate) . . . . . . . . . . . . . . . . . . . . . . 37
2.6.2.5 Effect of Monomer on Pyrolysis Products . . . . . . . . . . . . . 37
2.6.3 Catalytic Pyrolysis of HDPE . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.6.3.1 Zeolites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.6.3.2 Aluminosilicates . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.6.3.3 FCC Catalysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.6.3.4 Activated Carbon . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.6.3.5 Summary of Catalytic Cracking . . . . . . . . . . . . . . . . . . . 40
2.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
3 Development of Microwave-Assisted Pyrolysis Equipment . . . . . . . . . . . . . . 43
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.2 Description of the Apparatus as Received . . . . . . . . . . . . . . . . . . . . . 44
3.3 Maximising Transmission of Microwave Power to the Reactor . . . . . . . . . 45
3.3.1 Arc Prevention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.3.2 Minimising Energy Lost in Transmission to Reactor . . . . . . . . . . . 46
3.3.2.1 Locating Energy Loss Source . . . . . . . . . . . . . . . . . . . . 46
3.3.2.2 Simplifying the Power Transmission System . . . . . . . . . . . 46
3.3.3 Window Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.3.3.1 Miniaturisation: Integration of Water Cooling and Gas Inlet . 48
3.3.3.2 Investigating Alternative Window Materials . . . . . . . . . . . 48
3.3.3.3 Reduced Carbon Deposition from Silicone Material . . . . . . 49
3.3.3.4 Final Window Design Refinements: Added Strength from
Central Mica Layer . . . . . . . . . . . . . . . . . . . . . . . . . . 50
3.3.4 Maximising Power Transfer through Impedance Matching . . . . . . . 50
3.3.4.1 Background: Impedance Matching Theory . . . . . . . . . . . . 50
3.3.4.2 Tuning the Microwave Circuit . . . . . . . . . . . . . . . . . . . . 52
3.3.4.3 Time-Domain Impedance Optimisation . . . . . . . . . . . . . . 52
3.4 Sample Injection System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
3.4.1 Plunger Injection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
3.4.1.1 Problems with Plunger System . . . . . . . . . . . . . . . . . . . 55
3.4.2 Continuous Rotary Feeder . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
3.4.2.1 Calibration of Rotary Feeder . . . . . . . . . . . . . . . . . . . . 57
3.4.2.2 Co-addition of Air with HDPE Pellets . . . . . . . . . . . . . . . 58
3.4.2.3 Pellet Adherence to Reactor Inlet . . . . . . . . . . . . . . . . . 58
3.5 Pyrolysis Product Condensing Apparatus . . . . . . . . . . . . . . . . . . . . . 58
3.5.1 Liebig Condensers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
3.5.1.1 Multiple Condensers . . . . . . . . . . . . . . . . . . . . . . . . . 59
3.5.1.2 Temperature Variation . . . . . . . . . . . . . . . . . . . . . . . . 59
3.5.1.3 Attempts to Collect Aerosol Mist Using a Particle Trap . . . . 60
3.5.1.4 Difficulties with Sample Extraction . . . . . . . . . . . . . . . . 60
3.5.2 Alternative Condenser Designs . . . . . . . . . . . . . . . . . . . . . . . . 61
3.5.3 Dry Ice Cold Trap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
CONTENTS ix
3.5.4 Dry Ice Percolating Condenser . . . . . . . . . . . . . . . . . . . . . . . . 61
3.5.4.1 Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
3.5.4.2 CO2 Interaction with Condensate . . . . . . . . . . . . . . . . . . 63
3.5.4.3 CO2 Contamination in Collected Gas . . . . . . . . . . . . . . . . 64
3.6 Control Software Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3.6.1 Problems with Existing Control Setup . . . . . . . . . . . . . . . . . . . . 64
3.6.2 Hardware Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3.6.3 Software Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.6.3.1 Analog Input . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.6.3.2 PID Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
3.6.3.3 Practical Implementation Issues . . . . . . . . . . . . . . . . . . 68
3.6.3.4 Arc Prevention and Safety Systems . . . . . . . . . . . . . . . . 69
3.6.3.5 Data Logging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3.7 Safety Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3.7.1 Microwave Safety . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3.7.2 Preventing Equipment Leaks . . . . . . . . . . . . . . . . . . . . . . . . . 70
3.7.3 Extraction System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
3.7.4 Additional Safety Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
3.8 Final Configuration of Microwave-Assisted Pyrolysis Apparatus . . . . . . . . 71
3.9 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
4 Batch Pyrolysis of HDPE using an Activated Carbon Reactor Bed . . . . . . . . . . 75
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
4.2 Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
4.2.1 HDPE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
4.2.2 Coke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
4.2.3 Activated Carbon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
4.3 Microwave Pyrolysis Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
4.3.1 Experimental Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
4.3.2 Pyrolysis Product Mass Measurement . . . . . . . . . . . . . . . . . . . . 79
4.4 GC/MS Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
4.4.1 Condensed Product Analysis . . . . . . . . . . . . . . . . . . . . . . . . . 80
4.4.2 Gas Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
4.4.3 Quantification of Compounds by Peak Area . . . . . . . . . . . . . . . . 80
4.5 Analysis of Condensed Pyrolysis Oils and Waxes . . . . . . . . . . . . . . . . . 81
4.5.1 Visual Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
4.5.2 GC/MS Analysis and Peak Identification . . . . . . . . . . . . . . . . . . 81
4.5.3 Mass Distribution of Condensed Pyrolysis Products . . . . . . . . . . . 83
4.5.3.1 Operation at the Temperature Threshold for Pyrolysis . . . . . 85
4.5.3.2 Influence of Reactor Temperature on Cracking Rate and
Residence Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
4.5.3.3 Influence of Reactor Temperature on Molecular Mass Distri-
bution of Products . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
4.5.3.4 Bimodal Molecular Mass Distribution with Coke Reactor Bed 87
4.5.3.5 Production of Liquid Transport Fuels . . . . . . . . . . . . . . . 88
4.5.4 Chemical Composition of Condensed Pyrolysis Products . . . . . . . . 89
4.6 Analysis of Non-condensable Pyrolysis Gases . . . . . . . . . . . . . . . . . . . 92
4.6.1 Mass Distribution of Non-condensable Pyrolysis Gases . . . . . . . . . 92
4.6.2 Chemical Composition of Non-condensable Pyrolysis Gases . . . . . . 93
4.7 Conservation of Mass and Hydrogen Transfer . . . . . . . . . . . . . . . . . . 95
4.8 Pyrolysis Product Component Yields . . . . . . . . . . . . . . . . . . . . . . . . 96
4.8.1 Comparison with Other Published Work . . . . . . . . . . . . . . . . . . 97
4.8.1.1 Temperature Dependent Gas Production and Influence of
Reactor Configuration . . . . . . . . . . . . . . . . . . . . . . . . 99
4.8.1.2 Minimal Retained Portion . . . . . . . . . . . . . . . . . . . . . . 99
4.8.1.3 Relative Gaseous and Condensed Product Yields and the
Influence of Reactor Configuration used in This Work . . . . . 100
4.8.1.4 Comparison to Other Activated Carbon Work . . . . . . . . . . 101
x CONTENTS
4.9 Utility of Products Produced Using an Activated Carbon Bed . . . . . . . . . 102
4.9.1 Condensed Pyrolysis Oil . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
4.9.1.1 Aromatic Content . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
4.9.2 Gases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
4.10 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
5 Continuous Pyrolysis of HDPE using an Activated Carbon Reactor Bed . . . . . . 107
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
5.2 Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
5.3 Pyrolysis Product Component Yields . . . . . . . . . . . . . . . . . . . . . . . . 109
5.3.1 Coke Bed Control Yields . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
5.3.2 Activated Carbon Yields . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
5.3.2.1 Retained Fraction . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
5.3.2.2 Condensed Products . . . . . . . . . . . . . . . . . . . . . . . . . 110
5.3.2.3 Non-condensable Gases . . . . . . . . . . . . . . . . . . . . . . . 112
5.3.2.4 Change in Activity of Bed . . . . . . . . . . . . . . . . . . . . . . 112
5.4 Analysis of Condensed Pyrolysis Oils and Waxes . . . . . . . . . . . . . . . . . 113
5.4.1 Mass Distribution of Condensed Pyrolysis Products . . . . . . . . . . . 113
5.4.1.1 Coke Bed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
5.4.1.2 Activated Carbon Bed . . . . . . . . . . . . . . . . . . . . . . . . . 115
5.4.1.3 Temperature Dependent Change in Activated Carbon Bed
Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
5.4.1.4 Change in Carbon Number Distribution . . . . . . . . . . . . . . 117
5.4.2 Chemical Composition of Condensed Pyrolysis Products . . . . . . . . 119
5.4.2.1 Coke Bed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
5.4.2.2 Activated Carbon Bed . . . . . . . . . . . . . . . . . . . . . . . . . 120
5.4.2.3 Specific Compound Analysis and Examples . . . . . . . . . . . . 122
5.4.3 Economic Value and Consistency Over Time of Condensed Products . 125
5.5 Analysis of Non-condensable Pyrolysis Gases . . . . . . . . . . . . . . . . . . . 127
5.5.1 Chemical Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
5.5.1.1 Coke Bed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
5.5.1.2 Activated Carbon Bed . . . . . . . . . . . . . . . . . . . . . . . . . 127
5.5.2 Economic Value of Non-Condensable Gases . . . . . . . . . . . . . . . . 130
5.6 Temperature Dependence of Rate of Change of Catalytic Activity . . . . . . . 130
5.7 Batch and Continuous Operation Comparison . . . . . . . . . . . . . . . . . . . 131
5.7.1 Coke Bed Yields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
5.7.2 Chemical Composition of Products Produced using Coke Bed . . . . . 131
5.7.2.1 Molecular Mass Distribution and Heavy Aliphatic Grouping . 132
5.7.3 Residence Time Differences due to Pyrolysis Rate and Exit Pressure . 132
5.7.3.1 Simplified Mathematical Model of Input Modes . . . . . . . . . 133
5.7.3.2 Effect of Residence Time on Chemical Composition of Pyro-
lysis Products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
5.7.3.3 Convective Transfer of Pyrolysis Aerosol Particles Owing to
Exit Pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
5.7.4 Differences in Reaction Mode for Activated Carbon Studies . . . . . . 136
5.7.4.1 Deactivation Across Batch Experimental Set . . . . . . . . . . . 136
5.7.4.2 Yield and Composition Differences . . . . . . . . . . . . . . . . . 137
5.8 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
6 Mechanisms of Pyrolysis, Catalysis, and Deactivation . . . . . . . . . . . . . . . . . 141
6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
6.2 Pyrolysis Reaction Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
6.2.1 Thermal Decomposition of HDPE . . . . . . . . . . . . . . . . . . . . . . . 142
6.2.1.1 Initiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
6.2.1.2 Inter- and Intra-molecular Hydrogen Transfer . . . . . . . . . . 143
6.2.1.3 Beta-scission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
6.2.1.4 Termination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
6.2.1.5 Non-Catalytic Pyrolysis of HDPE . . . . . . . . . . . . . . . . . . 145
6.2.2 Cyclic and Aromatic Formation . . . . . . . . . . . . . . . . . . . . . . . . 145
6.2.2.1 Diels-Alder Reactions . . . . . . . . . . . . . . . . . . . . . . . . . 145
CONTENTS xi
6.2.2.2 Radical-Mediated Aromatic Formation . . . . . . . . . . . . . . 147
6.3 Catalytic Activity of Activated Carbon . . . . . . . . . . . . . . . . . . . . . . . 147
6.3.1 Catalytic Cracking Reaction Mechanisms . . . . . . . . . . . . . . . . . 148
6.3.1.1 Microporous Structure . . . . . . . . . . . . . . . . . . . . . . . . 148
6.3.1.2 High Energy “Defect” Sites Facilitating Hydrogen/Radical
Transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
6.3.1.3 Accounting for Activated Carbon Product Spectrum by Cata-
lytic Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
6.3.1.4 Oxygen Active Sites and Acid-Activated Catalysis . . . . . . . . 150
6.4 Microwave-Mediated Catalysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
6.4.1 Coupling to Free Radicals . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
6.4.2 Generation of Microplasmas through Microwave Irradiation . . . . . . 153
6.5 Changes in Porous Structure of Activated Carbon . . . . . . . . . . . . . . . . 155
6.5.1 Physical Adsorption Analysis Method . . . . . . . . . . . . . . . . . . . . 155
6.5.2 Adsorption/Desorption Isotherms . . . . . . . . . . . . . . . . . . . . . . 156
6.5.3 BET Surface Area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
6.5.3.1 Mathematical Accuracy of BET Model . . . . . . . . . . . . . . . 158
6.5.4 Kinetics of Deactivation/Surface Area Loss . . . . . . . . . . . . . . . . 158
6.5.4.1 Calculated Deactivation Over Batch Experiments . . . . . . . . 160
6.5.5 Distribution of Pore Volume . . . . . . . . . . . . . . . . . . . . . . . . . . 161
6.5.5.1 Temperature Dependence of Deposition Distribution . . . . . . 163
6.6 Chemical Composition of Mass Retained in Reactor . . . . . . . . . . . . . . 165
6.6.1 Thermogravimetric Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 165
6.6.1.1 Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
6.6.1.2 Thermogravimetric Analysis . . . . . . . . . . . . . . . . . . . . . 165
6.6.1.3 Infrared Spectrometry of Evolved Gases . . . . . . . . . . . . . 167
6.7 Electron Microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
6.7.1 Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
6.7.2 SEM Images . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
6.7.3 EDS Elemental Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
6.8 Attrition and Thermal Cycling . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
6.9 Mechanism of Activated Carbon Deactivation . . . . . . . . . . . . . . . . . . 171
6.9.1 Carbon Nanofilament Formation via Microwave-Induced Plasma . . . 174
6.10 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
7 Project Overview and Further Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
7.1 Project Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
7.1.1 Development of Microwave-Assisted Pyrolysis Equipment . . . . . . . 179
7.1.2 Exploratory Experiments with Batch Addition of HDPE . . . . . . . . . 180
7.1.3 Continuous Throughput Pyrolysis of HDPE . . . . . . . . . . . . . . . . . 181
7.1.4 Mechanisms of Pyrolysis, Catalysis, and Deactivation . . . . . . . . . . 183
7.2 Further Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
7.2.1 Isolation of Microwave Effects . . . . . . . . . . . . . . . . . . . . . . . . 185
7.2.2 Other Catalysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
7.2.3 Composition of Deposited Carbon Material . . . . . . . . . . . . . . . . 186
7.2.4 Regeneration of Activated Carbon . . . . . . . . . . . . . . . . . . . . . . 186
7.2.4.1 Preliminary Reactivation Data . . . . . . . . . . . . . . . . . . . 187
7.2.4.2 Future Reactivation Work . . . . . . . . . . . . . . . . . . . . . . 187
7.2.5 Modelling of the Pyrolysis Process . . . . . . . . . . . . . . . . . . . . . . 188
7.2.6 Future Reactor/Condensing Systems . . . . . . . . . . . . . . . . . . . . 189
7.2.6.1 Fractionating Column . . . . . . . . . . . . . . . . . . . . . . . . . 189
7.2.6.2 Optimising for Gaseous Hydrocarbon Production . . . . . . . . 191
7.2.7 Contamination and Diversification of Input Streams . . . . . . . . . . . 192
7.2.8 Energy Return on Investment . . . . . . . . . . . . . . . . . . . . . . . . . 192
7.2.9 Economic Analysis of the Process . . . . . . . . . . . . . . . . . . . . . . 193
7.3 Conclusions and Evaluation of Study Objectives . . . . . . . . . . . . . . . . . 194
7.3.1 Additional Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
7.4 Closing Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198

List of Figures
1.1 Cost breakdown of liquid transport fuels in the U.K. . . . . . . . . . . . . . . 10
1.2 Ethene monomer and its corresponding polyethylene polymer form . . . . . 11
2.1 Allotropes of carbon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.2 Schematic representation of pores in an activated carbon particle . . . . . 18
2.3 Schematic structure of activated carbon, showing the folds in the planar
structure responsible for porosity . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.4 Chemical structures of some plastics . . . . . . . . . . . . . . . . . . . . . . . 35
3.1 Microwave-assisted pyrolysis apparatus as received . . . . . . . . . . . . . . 44
3.2 Final design of the microwave-transparent window . . . . . . . . . . . . . . . 49
3.3 Simplified circuit used to illustrate the impedance matching principle . . . 51
3.4 Power transferred to the load as a function of the load impedance . . . . . 52
3.5 Microwave energy reflected from the reactor . . . . . . . . . . . . . . . . . . 53
3.6 Crests in the carbon bed causing periodic variation in load impedance . . . 54
3.7 Batch addition plunger injection apparatus . . . . . . . . . . . . . . . . . . . . 55
3.8 Drop in reactor temperature due to sample addition . . . . . . . . . . . . . . 56
3.9 Calibration of the continuous HDPE pellet rotary feeder . . . . . . . . . . . . 57
3.10 Configuration of different tested condenser designs . . . . . . . . . . . . . . 59
3.11 Final design of dry ice percolating condenser . . . . . . . . . . . . . . . . . . 62
3.12 Simplified PID controller software schematic . . . . . . . . . . . . . . . . . . 66
3.13 Screenshot of the final PID controller software interface . . . . . . . . . . . 67
3.14 Schematic of the final configuration of the microwave pyrolysis apparatus 72
3.15 Final configuration of the microwave pyrolysis apparatus . . . . . . . . . . . 73
4.1 Photo of condensed pyrolysis products using batch input of sample . . . . . 81
4.2 Typical GC/MS chromatograms generated from GC/MS analysis of con-
densate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
4.3 Distribution of compounds in condensed pyrolysis oil by number of
carbons in molecule from batch pyrolysis of HDPE . . . . . . . . . . . . . . . 84
4.4 Chemical composition of condensed pyrolysis products produced with
batch introduction of sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
4.5 Merging of aliphatic gas chromatograph peaks at larger molecular masses 90
4.6 Hydrogen to carbon proportion of unbranched hydrocarbons by carbon
number . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
5.1 Yields of retained, condensed, and non-condensable pyrolysis products
produced in continuous mode with an activated carbon bed . . . . . . . . . 111
5.2 Average molecular mass and range of condensed pyrolysis products with
continuous introduction of HDPE . . . . . . . . . . . . . . . . . . . . . . . . . . 114
5.3 Change in carbon number distribution of condensed products due to
deactivation of activated carbon . . . . . . . . . . . . . . . . . . . . . . . . . . 118
5.4 Chemical composition of the condensed pyrolysis products produced
using a coke bed with a continuous input of HDPE . . . . . . . . . . . . . . . 119
5.5 Change in chemical composition of the condensed pyrolysis products
produced using an activated carbon bed at 450°C with a continuous input
of HDPE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
xiii
xiv LIST OF FIGURES
5.6 Influence of reactor temperature on change in chemical composition of
the condensed pyrolysis products produced using an activated carbon bed
with a continuous input of HDPE . . . . . . . . . . . . . . . . . . . . . . . . . . 121
5.7 Proportion of condensate produced using an activated carbon bed with a
carbon number in the liquid transport fuel range (C≤21) . . . . . . . . . . . . 126
5.8 Theoretically derived pyrolysis rates in batch and continuous modes at
high and low reaction temperatures . . . . . . . . . . . . . . . . . . . . . . . . 134
6.1 The formation of aromatic species through the Diels-Alder reaction . . . . . 146
6.2 Oxygen-based functional groups present on the surface of activated carbon151
6.3 Microplasmas generated in the activated carbon bed induced by micro-
wave irradiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
6.4 Nitrogen adsorption/desorption isotherms from activated carbon and coke
reactor beds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
6.5 Change in surface area of the activated carbon as it processes progress-
ively more HDPE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
6.6 Arrhenius plot to determine activation energy of deactivation process . . . 159
6.7 Calculated change in BET surface area throughout batch experiments . . . 160
6.8 Cumulative pore volume of activated carbon and coke reactor beds, as
pore diameter decreases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
6.9 Temperature dependence of coke deposition, and effect on surface area
of activated carbon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
6.10 Influence of reaction rate and diffusion rate of pyrolysis gas on spatial
distribution of coke deposition . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
6.11 Thermogravimetric curves for virgin and used activated carbon . . . . . . . 166
6.12 Typical FT-IR spectrum of evolved gases during TGA of carbon samples . . 167
6.13 SEM micrographs of activated carbon . . . . . . . . . . . . . . . . . . . . . . . 169
6.14 Typical energy-dispersive X-ray spectrum of activated carbon sample . . . 170
6.15 Testing for attrition by change in bed mass and surface area over time . . . 171
6.16 Carbon filament formation resulting from the creation of microplasmas in
the hydrocarbon pyrolysis gas . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
7.1 Activated carbon steam reactivation at 700°C using existing microwave
reactor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
7.2 Schematic of fractionating column reactor . . . . . . . . . . . . . . . . . . . . 190
List of Tables
1.1 Common hydrocarbon products by carbon number . . . . . . . . . . . . . . . 9
2.1 Reactions catalysed by activated carbon . . . . . . . . . . . . . . . . . . . . . . 22
3.1 Input and output requirements for computer control of the microwave
power supply . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
4.1 Typical material properties of Liten ML 71 HDPE . . . . . . . . . . . . . . . . . 77
4.2 Typical material properties of coke and activated carbon used in this work . 77
4.3 Comparison of the molecular weight of condensed pyrolysis products
produced in batch mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
4.4 Molecular weight of non-condensable pyrolysis products produced with
batch introduction of sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
4.5 Chemical composition of non-condensable pyrolysis gases produced with
batch introduction of sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
4.6 Yields of retained, condensed, and non-condensable pyrolysis products
produced with batch introduction of sample . . . . . . . . . . . . . . . . . . . . 97
4.7 Yields of polyethylene pyrolysis reported in the literature, operating in
batch mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
5.1 Yields of retained, condensed, and non-condensable pyrolysis products
produced in continuous mode with a coke bed . . . . . . . . . . . . . . . . . . . 109
5.2 Five most prevalent compounds for each chemical grouping present in the
condensate produced using a coke bed continuously fed with HDPE . . . . . 123
5.3 Five most prevalent compounds for each chemical grouping present in the
condensate produced using an activated carbon bed continuously fed with
HDPE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
5.4 Chemical composition of the non-condensable gases produced using a coke
bed using a continuous input of HDPE . . . . . . . . . . . . . . . . . . . . . . . 128
5.5 Composition of non-condensable gases produced using an activated carbon
bed using a continuous input of HDPE . . . . . . . . . . . . . . . . . . . . . . . 129
5.6 Comparison between condensed product properties produced in batch and
continuous reactor modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
xv

Nomenclature
Roman
E Error from target temperature [°C]
k Constant
P Power [W]
T Temperature [°C]
t Time [s]
V Voltage [V]
Z Electrical Impedance [Ω]
Subscripts
0 Initial
D Derivative
I Integral
L Load
P Proportional
S Source
t time step
Abbreviations
CN, Cx the number of carbons in a molecule, where x is the
number of carbons
EDS Energy-dispersive X-ray spectroscopy
FBR Fluidised bed reactor
FCC Fluid cracking catalyst/catalysis
FT-IR Fourier transform infrared spectroscopy
GC/MS Gas chromatography-Mass spectrometry
HDPE High density polyethylene
LDPE Low density polyethylene
MW Molecular weight [g mol-1]
PAH Polyaromatic Hydrocarbon
PID Proportional-Integral-Derivative
PMMA Poly(methylmethacrylate)
PP Polypropylene
PS Polystyrene
PTFE Polytetrafluoroethylene, Teflon
PVC Poly(vinyl chloride)
SD Standard deviation
SEM Scanning electron microscopy
wt% Weight percentage
xvii

1
Introduction
1.1 The Problem and Opportunity of Plastic Waste
For a material with a relatively recent genesis, plastics have certainly come to
play a dominant role in modern goods and manufacturing. Their low cost, ease of
manufacture, and versatility, have led to an enormous growth in their consumption
(of roughly 5% a year over the last 20 years), with 265 million tonnes produced
worldwide in 2010, of which 57 million tonnes are accounted for by Europe alone
[1], and this growth is only expected to continue. This surfeit of plastics presents
a significant disposal issue, especially considering the large quantities of single-use
items and packaging materials that are fabricated from plastics, and the fact that
as highly inert materials, most plastics degrade in the environment only over very
long periods of time. Furthermore, the production of plastics represents a significant
energy investment, in the extraction and refinement of petrochemical feedstock, and
the subsequent synthesis of polymer. As a high-energy material, at the end of a plastic
item’s useful life, there is significant potential for resource recovery, either to re-use the
1
2 Introduction
plastic itself, or to recover some of the stored energy that has gone into its manufacture.
In any case, it is a lamentable waste simply to “throw away” plastic after one use cycle,
and with it all the latent energy inherent to the material, as happens all too often in
current practice.
1.2 Current Plastics Recycling Techniques
The U.K. sees relatively low rates of plastics recycling, with 69% of post-consumer
plastic waste generated in 2010 ending up in landfill (compared with 42% across the
whole of Europe), with no attempt at recycling or energy recovery made [1]. However,
the U.K. government has plans to double the rate of plastic packaging recycling (from
the 2007 figure of 24%) by the year 2020 [2], a commitment that will require significant
investment in additional recycling capability and techniques.
1.2.1 Mechanical Recycling
The simplest and most prevalent form of plastics recycling (comprising 24.3% of
plastics disposal in Europe [1]) is achieved by mechanical means: thermoplastics may
be melted and moulded into a new form, usually after being shredded or granulated.
However, different polymers generally do not mix well, necessitating separating plastic
types from one another before processing occurs — a time-consuming and costly
process. Indeed, even within a single type of polymer, small differences in composition
can cause phase separation when samples are melted together, resulting in structural
flaws and a lower quality final product. These issues are only exacerbated by the
different additives such as dyes and fillers that each manufacturer uses. Plastics must
also be contaminant free in order to be processed in this way, which can necessitate the
introduction of a costly washing step, depending on the state of the source material.
Polymer degradation over time can also result in inferior product produced using this
technique.
1.2.2 Energy Recovery
Typical plastics have a calorific value between 40–48 MJ/kg, greater than that of
coal [3, 4]; this high energy density makes plastics relatively sought after for energy
recovery purposes. Typically this is achieved through incineration, where the polymers
are exothermically oxidised in furnaces, and the resulting heat is used to generate
1.2 Current Plastics Recycling Techniques 3
electrical energy. This method can be used with all types of plastics; however, it suffers
several drawbacks. The combustion of plastics is essentially equivalent to burning
fossil fuels and results in the release of CO2 — a greenhouse gas that contributes to
climate change. Furthermore, incineration recovers only thermal energy from plastics,
and does not allow for the recovery of any of the chemical value of the polymers.
Incineration can also result in the release of toxic compounds (e.g., dioxins and furans)
that cause environmental damage and health problems [5, 6], as well as producing
hazardous compounds, such as fly ash, which are difficult to dispose of properly [7].
Currently incineration accounts for around 8% of plastics disposal in the U.K., though
the rates are much higher in western Europe as a whole, at 33.6% [1].
1.2.3 Feedstock Recycling
Feedstock recycling breaks down polymers into smaller molecules that can be used
as inputs for the creation of new chemicals and materials. This term incorporates a
range of processes such as chemical depolymerisation (yielding the original monomer),
thermal decomposition such as pyrolysis, and gasification, where hydrocarbons are
broken down into synthesis gas (H2 + CO). This gas can then be used as a fuel
directly, or converted into liquid hydrocarbon fuel through the Fischer-Tropsch process
[8]. While feedstock recycling is able to deal with a wide variety of plastics, and
is generally able to cope with fairly high levels of contamination, it usually requires
significant energy and capital intensive plants that only operate economically on large
scales [9]. The use of feedstock recycling techniques is not currently widespread, with
only 1.6% of plastic waste in Europe recycled using these methods at present [10].
Unfamiliarity and high capital cost barriers to entry are thought to contribute to the
current low prevalence of these techniques, though clearly this is a route with the
potential to expand in capacity significantly, especially given continuing technological
development, and the forecasted continuing rise in the cost of the petrochemicals that
these techniques produce, as easily accessible global supplies of raw materials are
depleted.
1.2.3.1 Pyrolysis
Pyrolysis is the thermal degradation of a substance in an inert environment (i.e.,
without the presence of oxygen), and may be used as a plastics feedstock recycling
technique where polymers are cracked into shorter hydrocarbons (their constituent
monomers, parts or multiples thereof, and more complex species produced when these
4 Introduction
products react together) — the final products resulting from this process comprise
waxes, oils, gases, and char. It is possible to optimise the process to maximise
production of any of these constituents by altering parameters such as process
temperature and reactant residence time [11]. For example, a high temperature and
residence time promotes gas production; high temperature and low residence time
(“flash pyrolysis”) maximises the yield of condensates; while a low temperature and
heating rate promotes char production [12]. The relatively wide variety of products
produced by the pyrolysis of plastics means that they usually have to be separated and
purified before use; fortunately this has the potential to be achieved on a commercial
scale with existing petrochemical distilleries and refineries.
1.2.3.2 Catalytic Pyrolysis
It is also possible to pyrolyse plastics in the presence of catalysts, which enhance
cracking and alter the yields of the pyrolysis products so that lighter products are
produced more abundantly. Most studies that have been carried out on catalytically
pyrolysing plastics use fluid cracking catalysts (FCCs) such as zeolites and alumino-
silicates, which have been developed by the petrochemical industry for the cracking of
long-chain hydrocarbons into lighter liquid fuels [13, 14].
1.3 Microwave-Assisted Pyrolysis
1.3.1 Microwave Heating
“Microwaves” are the portion of the electromagnetic spectrum between 300 MHz
and 3 GHz. The practical production and utilisation of this frequency band was
first carried out during World War Two in the development of high-definition radar.
During this time the cavity magnetron was invented — an inexpensive, highly efficient,
high-power microwave source that has allowed the application of microwaves for a
range of purposes, heating foremost amongst them, and that has directly enabled the
proliferation of the microwave oven so commonplace in households today. The heating
of a material by microwaves arises from the resistance of movement by dipoles and
charged species to the forces imposed by the alternating electromagnetic microwave
field [15].
1.3 Microwave-Assisted Pyrolysis 5
1.3.1.1 Advantages of Microwave Heating
Microwave heating offers several advantages over conventional techniques such as
electrical furnaces:
• As a non-contact form of heating, it is physically gentle, allowing for a diverse
range of applications over and above contact surface heating [16–18].
• It is a volumetric heating method, with the diffuse and pervasive nature of the
electromagnetic field able to heat many substances evenly and in bulk, without
relying on slower and less efficient conductive or convective techniques [19, 20].
• Microwave heating is able to achieve high power densities, allowing rapid heating
and the ability to reach high temperatures in short periods of time, which can
increase production speeds and reduce costs [21].
• Microwave fields interact only with particular types of materials, giving this form
of heating the ability to selectively target areas and substances (and not heat
surrounding air or containers). This can dramatically reduce the amount of
heating required to achieve a particular end, and with it the associated energy use
and costs. By selectively heating only targeted reactants it is possible to achieve
a more uniform temperature profile and a higher yield of desired end products
[22–24].
• Microwave heating is easily and rapidly initiated and terminated, especially
compared with processes that rely on combustion or indirect heating such as
steam.
• The electrical origin of microwave heating allows for direct electronic control,
providing a higher level of precision, safety, and automation.
• As microwave heating is an electromagnetic means of transmission it is possible
to delocalise the energy source from the heating target, with microwaves easily
transported over long distances through waveguides. This facilitates modular
construction of reactors with easier access and lower maintenance costs [25].
• Magnetrons are available in a wide range of sizes and power outputs, making
processes based on this form of heating highly scalable, enabling the same
technology used in laboratory-scale experimental reactors to be employed in
commercial sized plants, and potentially enabling smaller scale community sized
facilities to be constructed with lower initial capital costs.
6 Introduction
1.3.1.2 Disadvantages of Microwave Heating
Microwave heating also suffers from some drawbacks in comparison with conventional
heating methods:
• Microwave radiation presents an additional hazard over traditional heating meth-
ods, although this is easily contained within an appropriate Faraday cage. The non-
ionising nature of the radiation also limits risk in the event of exposure, with the
radiation having insufficient energy per quantum to create biologically damaging
ionised atoms or molecules.
• While the property of microwave heating only to target microwave-receptive
materials can be an advantage in some cases, in the event that microwave-
transparent materials (such as plastics) require heating, an intermediate micro-
wave receptor must be used.
• With the conversion from electrical energy to electromagnetic radiation to heat,
microwave heating is not 100% efficient, though modern equipment is able to
achieve over 90% conversion efficiency [26], and this loss is often more than
made up for by the ability to selectively target the area to be heated as previously
discussed.
• Using microwave heating places limits on the materials that can be used in
the construction of a reactor. While metal is essential in an electromagnetic
containment role, its use should be minimised within the reactor itself to avoid
causing arcing, with sharp metallic corners and small physical gaps between metal
surfaces presenting the highest risk of potential arc formation.
• The more complex apparatus required for microwave heating (i.e., magnetron,
power supply, waveguides, etc.) means that this form of heating costs more than
conventional methods. Furthermore, the cost of microwave heating apparatus
scales disproportionally with size, i.e., larger heating installations cost more per
watt than smaller ones, though this may be alleviated to some extent by using
multiple smaller units in parallel.
1.3.2 Microwave-Assisted Pyrolysis of Plastics using a Carbon
Bed
Microwave-assisted pyrolysis describes the application of microwave heating to a
substance to cause it to undergo pyrolytic breakdown (in the absence of oxygen). It is a
relatively new process, first developed by Tech-En Ltd (Essex, U.K.) [27, 28] to combine
1.3 Microwave-Assisted Pyrolysis 7
the advantages of microwave-mode heating with the commercial and environmental
benefits that the pyrolysis of plastics offers. Because of the microwave transparency of
plastics, in this application an intermediate receptor is used to absorb the microwave
energy and transfer it to the plastic via conduction.
1.3.2.1 Method for Microwave-Assisted Pyrolysis of Plastics Using a Carbon
Bed
The microwave-assisted pyrolysis of plastics is performed by heating a reactor bed to
a sufficiently high temperature, in the absence of oxygen, that added plastic is cracked
into smaller hydrocarbon molecules. The stochastic nature of the pyrolysis process
means that products with a wide range of molecular masses are usually produced.
These products exit the reactor in the gaseous state and are collected as condensed
waxes and oils, and as non-condensable light-chain gases. The cracking, volatilisation,
and physical transfer involved in this process also act as a separation mechanism, with
many forms of contaminant carbonising and remaining in the reactor, along with any
non-volatile species; as a result, microwave-assisted pyrolysis has a relatively robust
ability to cope with contaminants without significant impact on the final products [29,
30].
1.3.3 Carbon as a Reactor Bed Material
Particulate elemental carbon is the most common material used as a reactor bed in
microwave-assisted pyrolysis, owing to its excellent microwave receptivity and low
cost. Short transfer distances, the enveloping nature of a particulate bed, and small
particle size (with corresponding large surface area) make this a highly efficient method
of heat transfer, while by controlling the size of the carbon bed it is possible to localise
the heat dispersion within a small area, and maintain the targeted efficiency that
microwave heating offers. A further advantage of using carbon as a reaction bed
for pyrolysis is the highly reducing environment that it provides, which prevents the
formation of undesirable oxidised species, and again enhances the ability of the process
to cope with potential contaminants.
1.3.3.1 Coke
The “carbon” used in carbon beds for microwave-assisted pyrolysis is almost always
coke (examples of which are discussed in depth in §2.5). Coke is the solid carbonaceous
8 Introduction
residue that remains after heating a (primarily) hydrocarbon material to a high
temperature in the absence of oxygen (i.e., undergoing pyrolysis of some description
itself). It can be produced from a variety of sources, historically deriving from coal,
though is now a common by-product of the petroleum industry, produced during the
upgrading of heavy oil fractions to lighter hydrocarbons [31]. Owing to its high carbon
content, coke may be used as a fuel in its own right, or as a reducing agent in the
smelting of iron ore in blast furnaces [32, 33]. Its abundance as a common by-product
accounts for its low sourcing cost.
1.3.3.2 Activated Carbon
Activated carbon, the focus of the work presented in this dissertation, is a manufac-
tured form of carbon, with a central characteristic of an extremely high adsorptive
capacity that derives from microporous structure and resulting high surface area. It
is manufactured from various polymeric carbohydrates and hydrocarbons such as hard
woods, coconut shell, fruit stones, coals, and oil bitumen. These are first carbonised
(through pyrolysis), and then subjected to a chemical or thermal “activation” process
(discussed in depth in §2.3.2) that maximises the material’s porous potential, creating
and widening micropores on the molecular level, and introducing or enhancing high
energy active sites on to the surface.
The very high adsorptive ability of activated carbon has led to its current main
industrial use in purification (e.g., water treatment, decolourising). In the chemical
context, activated carbon sees high usage as a support material in heterogeneous
catalysis (with, e.g., platinum or palladium) to enhance effectiveness and minimise
cost [34, 35]. In the recent past, an appreciation has grown for the potential to use
activated carbon as a catalyst in its own right, with it being employed commercially
in the production of carbonyl chloride from carbon monoxide and chlorine, and it has
also been shown to be an effective catalyst for a variety of other reactions involving
hydrogen, oxygen, and halogens [36, 37] (a comprehensive account of the applications
of activated carbon is given in §2.3.3). More recently, activated carbon has shown
promise in the dehydrogenation of hydrocarbons [38, 39], suggesting it may also have
the potential to be employed in the catalytic cracking of long-chain polymers such as
plastics.
1.4 Petrochemicals and Hydrocarbon Fuels 9
Table 1.1: Common hydrocarbon products by carbon number [41, 42].
Product Carbon Number
Natural Gas C1
Ethene C2
Liquid Petroleum Gas (LPG) C3–C4
Petrol/Gasoline C5–C12
Kerosene/Jet Fuel C8–C16
Diesel C8–C21
Lubricating Oil C20–C40
Waxes >C40
HDPE Chundreds of thousands
1.4 Petrochemicals and Hydrocarbon Fuels
Petrochemicals and hydrocarbon fuels are central to current human civilisation, with
petrochemicals underpinning innumerable products such as plastics, fibres, solvents,
pharmaceuticals, and fertilizer, and oil-derived liquid and gas fuels making up 54.3% of
total primary energy supply worldwide [40]. That these chemicals have risen to such
prominence is a result of several factors, including:
• Historical abundance, with correspondingly cheap (in both monetary and energy
terms) extraction and refinement, in comparison with other energy sources.
• Network effects resulting from infrastructure development over time.
• Ease of transport, with the ability to be shipped long distances via pipeline or
tanker.
• Availability on a scale that matches societal demand.
• Advantageous physiochemical properties, with liquid transport fuels (e.g., petrol
and diesel) exhibiting unparalleled volumetric energy density (excepting nuclear
power).
Table 1.1 lists a selected number of common hydrocarbon products by carbon number.
The point developed here is that these substances share a common chemical heritage
and structure: they are almost always derived from refined crude oil, and are made up
of hydrocarbons of various molecular masses, distinguished from one another simply
by the number of carbon-hydrogen units present in the molecule. As a result of this
underlying similarity of chemistry, it is possible to convert between pure hydrocarbon
products by adding units together (e.g., polymerisation and Fischer–Tropsch synthesis)
or breaking them apart, achieved through pyrolysis and cracking.
10 Introduction
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Petrol Diesel
N
e
t 
C
o
s
t 
Distribution + Profit
Refining
Cost of Oil
Figure 1.1: Cost breakdown of liquid transport fuels in the U.K. Cost of oil includes
exploration/discovery and extraction/production. Distribution + Profit includes retail margin,
supply, and distribution. This figure presents net costs and as such excludes duty and VAT,
though these contribute significantly to the final cost paid by the consumer. Data adapted from
[43].
1.4.1 Cost Savings and Efficiency Benefits of Feedstock
Recycling
Figure 1.1 gives a breakdown of the net cost of liquid transport fuels in the U.K. By far
the greatest proportion of the cost of the final refined product is the cost of the crude oil
from which the fuels are made (at 84–89%), which includes exploration/discovery and
extraction/production. Refining, distribution, and retail profit make up surprisingly
little of the overall costs. By repurposing a waste stream into a raw material from
which new products can be fabricated, but one in which the significant discovery
and extraction costs have already been undertaken, microwave-assisted pyrolysis has
enormous potential to make savings in both monetary terms [43] and in terms of the
energy required to produce the next cycle of products [44].
1.5 High Density Polyethylene (HDPE) 11
C
H
H
C
H
H
Ethene
C
H
H
C
H
H
Polyethylene


n
Figure 1.2: Ethene monomer and its corresponding polyethylene polymer form.
1.5 High Density Polyethylene (HDPE)
While the focus of this introduction thus far has been on plastics generally, this
dissertation focuses on high density polyethylene (HDPE) specifically, using it as a
model for more general hydrocarbon polymers. Polyethylene is the most common
form of plastic, accounting for 29% of total demand in Europe [1]. HDPE is made
through chain-growth polymerisation, combining ethene molecules together into chains
between 500,000–1,000,000 units long by the application of heat and pressure [14]
(see figure 1.2). The “high density” in HDPE describes polyethylene produced with
little structural branching, which is ensured by the use of an appropriate catalyst (such
as the titanium-based Ziegler–Natta class [45]). This linearity causes HDPE to have a
greater proportion of crystalline regions than “low density” polyethylene, and it is as
a result harder, and exhibits a higher tensile strength [14]. These properties result in
HDPE being employed in a diverse set of applications including bottles and containers,
plastic bags, and a wide variety of consumer goods.
HDPE was chosen as a model material for this research because of its abundance
and widespread use, the simplicity of its pure hydrocarbon molecular structure (thus
avoiding adding complication to the already complex pyrolytic breakdown reaction
pathways), and because it is one of the relatively few materials on which microwave
pyrolysis research has already been conducted [29, 46], thus offering an external point
of comparison and validation to this study.
1.6 Potential Benefits from the Microwave-Assisted
Pyrolysis of HDPE
In summary, a number of potential benefits arise from the microwave-assisted pyrolysis
of HDPE:
• By presenting an alternative disposal route for plastics, microwave-assisted
pyrolysis has the potential to divert this waste stream away from incineration
12 Introduction
(with its associated toxic emissions and greenhouse gas release) and landfilling
(for which space is at a premium in many countries, and which in any case results
in extremely long decomposition times).
• With the potential to cope with higher levels of contamination than conventional
recycling techniques, microwave-assisted pyrolysis presents a pathway to recycle
waste streams that would otherwise be disposed of through less efficient means
and/or without resource recovery.
• By creating a valuable end product from a waste stream, microwave-assisted
pyrolysis assigns value to that waste, which motivates collection and disposal
through this value-added pathway, and potentially disincentivises illegal and
inappropriate disposal, which currently causes significant environmental damage
[47–49].
• New plastic created from the raw materials produced by the microwave-assisted
pyrolysis of waste plastic should be indistinguishable from virgin plastic produced
from oil-derived feedstock, avoiding the pitfalls and lower quality inherent to
polymer produced through contemporary mechanical recycling techniques.
• While any hydrocarbon fuels produced through the microwave pyrolysis of plastic
would contribute equally to greenhouse gas emissions as those produced through
conventional means if their combustion alone is considered, given the potential
energy savings gained by avoiding the need to discover and extract fresh raw
materials for their production (with the CO2 emissions associated with that
process), the overall energy use and emissions for fuels produced by microwave-
assisted pyrolysis should be less. This makes fuel produced in this manner far
preferable, given the fungibility of the international hydrocarbon market, and
constant or increasing demand.
• Non-fuel petrochemicals, such as virgin plastic, produced using the products of
microwave-assisted pyrolysis as feedstock, should have significantly lower energy
and greenhouse gas emissions associated with their manufacture than those
derived from conventional sources.
1.7 Motivation and Objectives for the Study
It is clear that microwave-assisted pyrolysis has a range of potential environmental,
economic, and quality benefits as a plastic recycling technique; however, the stochastic
nature of the pyrolytic degradation of hydrocarbon plastics results in the production of
a wide range of molecular masses, which necessitates costly processing and separation
1.7 Motivation and Objectives for the Study 13
of the desirable short-chain hydrocarbons from the less useful long-chain waxes.
Catalytic pyrolysis has the potential to narrow the range of molecules produced;
microwave pyrolysis requires the use of a carbon bed to transfer microwave energy
to microwave-transparent plastics; activated carbon has the potential to combine these
two concepts by using the catalyst as the microwave-absorbing intermediate.
Generally, the microwave-pyrolysis field is inchoate but burgeoning, as is the use of
activated carbon as a catalyst in its own right. While a small body of work exists on the
microwave-assisted pyrolysis of HDPE [29, 46], this is the first such application using
activated carbon as a catalytic reactor bed. Accordingly, a significant portion of this
study is devoted to fundamental work on developing and refining necessary apparatus
and demonstrating feasibility. Indeed, a significant motivation behind this work is basic
curiosity to “see if it will work”.
A second motivation underlying this study is to develop a process that will facilitate
an improvement in the efficiency of resource use. Instead of throwing away material
that has a large amount of energy invested in it, if the technology exists and is feasible,
it makes sense to make the best use of already invested energy by maximising the
utility of said material, opening up the possibility of a second use cycle, and avoiding
expending energy to extract and process increasingly scarce new raw material from
the surface of the earth.
In light of the above, the objectives for this study are as follows:
• To develop suitable apparatus to reliably carry out the microwave-assisted pyro-
lysis of HDPE, and collect the resulting pyrolysis products.
• To determine whether or not activated carbon is able to act as a catalyst in the
microwave-assisted pyrolysis of HDPE.
• To characterise the products produced by the microwave-assisted pyrolysis of
HDPE, and evaluate their potential utility.
• If initial results are sufficiently promising and further investigation is warranted,
to examine the scale-up of the process with a larger capacity and continuous
throughput.
• To understand and characterise the catalytic properties of activated carbon in the
context of the pyrolytic breakdown of HDPE.
• To evaluate the potential of microwave-assisted pyrolysis of HDPE for future
commercial development on an industrial scale.
14 Introduction
1.8 Outline of Thesis
Having introduced the foundational concepts and motivations underpinning this study
in this chapter, chapter 2 presents a review of previous work in the literature carried
out by other researchers relevant to the current study. The development of the
microwave-assisted pyrolysis apparatus and the condensation system used for product
collection is described in chapter 3. This development allowed an initial feasibility
study to be carried out (chapter 4) that investigated the use of an activated carbon
bed to conduct microwave-assisted pyrolysis of HDPE, with small quantities of HDPE
introduced to the reactor in a batch fashion at a range of reactor temperatures, and an
analysis of the products that resulted. On the strength of these results, a rotary feeding
mechanism was developed that allowed the processing of larger quantities of plastic
with the continuous introduction of material. Chapter 5 assesses the products resulting
from the microwave-assisted pyrolysis of HDPE in this reaction mode, and details the
changes in the process, products, and reactor bed observed with the processing of
larger quantities of material over time. Chapter 6 examines the potential reaction
mechanisms that account for the pyrolytic breakdown of HDPE, and the additional
pathways that may occur as a result of the activated carbon catalyst. The physical
changes in the reactor bed, and in its catalytic ability, as the bed is used to process
increasing amounts of material, are also discussed in this section. Finally, chapter 7
provides an overview and candid appraisal of the work carried out in this study, and
describes potential avenues of further study that could be carried out to advance this
field further.
2
Literature Review
2.1 Carbon
Carbon is the fourth most abundant element in the universe [50], and perhaps the
most versatile; it is non-metallic, and tetravalent — with four electrons available to
form covalent bonds with other atoms in sp3 (tetrahedral), sp2 (ring), and sp (linear)
configurations. Its molecular properties and bonding abilities allow carbon to form
more compounds than any other element, including the entire gamut of organic
compounds, which form the chemical basis of all known life. Carbon also has the ability
to form long chains of interconnecting C-C bonds — a property known as catenation,
and which is the basis of the element’s ability to form macromolecules and polymers
such as DNA, proteins, carbohydrates, and lipids, as well as synthetic polymers such
as rubbers and plastics — the focus of this study. Carbon is also used in this study as
a microwave receptor in the form of coke, and microwave receptor and catalyst in the
form of activated carbon, purposes for which it has physical and chemical properties
that make it particularly useful, as is developed throughout this review.
15
16 Literature Review
Figure 2.1: Some allotropes of carbon: a) diamond, b) graphite, c) lonsdaleite, d–f) fullerenes
(C60, C540, C70), g) glass-like carbon, h) carbon nanotubes. Reproduced from [51] under creative
commons licensing.
2.1.1 Allotropes of Carbon
As a result of its versatile bonding abilities, elemental carbon is able to form a variety of
physical molecular forms, or allotropes, the most well known of which include diamond,
graphite, and polycrystalline carbon (colloquially, though not technically, “amorphous”)
such as coal and soot. Diamond (figure 2.1a) is an extremely hard form of carbon
where each carbon atom is bonded to four others in a tetrahedral lattice (its rarer
cousin lonsdaleite (figure 2.1c) employs a hexagonal lattice structure but has similar
properties).
Graphite (figure 2.1b) is the most thermodynamically stable form of carbon, with
a hexagonal lattice structure that forms planar layers; an individual hexagonal plane
is known as a graphene layer. With each carbon atom bonded to three others, the
remaining valence electrons delocalise, which allows graphite to conduct electricity.
2.1 Carbon 17
Adjacent layers are stacked atop and displaced from one another to form an alternating
ABAB sequence; these are held together by van der Waals forces and resonance effects
such that the force between layers is approximately one third that of a carbon-carbon
double bond [52].
Apart from diamond and lonsdaleite, all known carbon forms are related to the
graphite lattice structure, with a continuous spectrum of increasing disorder existing,
where layers abandon the ABAB stacking, and become smaller and more defective, until
eventually, as in the highly disordered glass-like carbons (figure 2.1g), the existence of
identifiable graphene layers is almost lost. The more exotic allotropes of carbon such
as fullerene spheres (figure 2.1d–f) and nanotubes (figure 2.1h), while clearly on the
more ordered end of this spectrum, lie outside the purview of this study.
The two forms of elemental carbon dealt with in this study are coke, and activated
carbon. Coke is the carbonaceous residue that remains after the high temperature
pyrolysis of a hydrocarbon material (distinct from ash, which is the residue that remains
after a substance is burned, and which consists primarily of metal oxides). Activated
carbon is described in detail in §2.3.
2.1.2 Non-Graphitisable Carbon
Research carried out by Rosalind Franklin (a pioneer in research into the structure
of coals and carbons as well as her perhaps more widely known work on DNA)
in the 1950s has led to the classification of carbon compounds into graphitisable
(which are materially anisotropic) and non-graphitisable (isotropic) categories, based
on their ability to adopt the long-range parallel planar structure (with associated X-
ray crystallography patterns) that is characteristic of graphite, when they are heated
to over 2000°C [53, 54]. While non-graphitisable carbon exhibits many localised
elements of graphene lattice structure (hence its description as polycrystalline), it
does not exhibit parallelism on a long-range basis; instead, non-planar elements result
in complex three-dimensional structures. The inability of non-graphitisable carbon to
change conformation arises from the short range structure of the non-planar units,
which are bonded into the overall structure during carbonisation; the activation energy
required to simultaneously break so many bonds in order to adopt a graphite structure
is prohibitive in all but the most extreme cases [52]. Activated carbon is such a non-
graphitisable substance, and it is from this characteristic that its porosity and other
useful properties derive.
18 Literature Review
Macropores
>50 nm
Mesopores
2–50 nm
Micropores
< 2nm
Figure 2.2: Schematic representation of pores in an activated carbon particle. Pores are able
to extend deep into a substance in three dimensions, significantly increasing its overall surface
area.
2.2 Porosity
Porosity is quite simply empty space within a structure into which substances of equal
or smaller size may adsorb. As illustrated in figure 2.2, porous fissures can extend
deep within a substance’s structure, significantly enhancing its overall surface area.
Porosity is usually categorised into three classes: macropores, with a diameter of
>50 nm; mesopores which measure 2.0–50 nm, and micropores with a diameter of
<2.0 nm. While small molecule adsorption mostly takes place in micropores (these
being of comparable size magnitudes), macro- and mesopores play an important role in
providing passage for adsorbate to reach the micropores.
2.3 Activated Carbon
Activated carbon is a form of manufactured, non-graphitisable carbon with a very high
porosity and absorptive ability, and which shows promising catalytic activity (discussed
later in this section) that has made it the subject of this study.
2.3.1 Material Origin and Carbonisation
Activated carbon can be manufactured from a wide range of macromolecular carbo-
hydrates and hydrocarbons of natural and synthetic origin, the main requirement of
the source material being that it is primarily carbon containing. Activated carbon has
2.3 Activated Carbon 19
Figure 2.3: Schematic structure of activated carbon, showing the folds in the planar structure,
and defective graphene lattices, from which the microporous structure of activated carbon
derives. Adapted from [71].
been successfully produced from wood [55–57], coconut shell [58–60], fruit stones and
nut shells [61–65], coals [66–68], and oil bitumen [68, 69].
These input materials are carbonised/pyrolysed — treated at high temperature in
an inert atmosphere, which results in an increase in carbon percentage as volatile
gases and liquids are driven off and the number of non-carbon heteroatoms decreases.
Research has shown that throughout this process, the macromolecular structure of
the material remains mostly intact [70], though carbon atoms rearrange over small
distances (probably <1 nm) within the network to positions of greater stability, i.e.,
closer to the six-membered graphene ring structure [52].
The carbon material remaining at the end of this process will have significant
porous structure, with carbon atoms arranged in graphene sheets crosslinked in a
random manner. Defects in the graphene lattice cause complex three-dimensional
structure including folds and curves in the sheet, and spatial distortions of the sheet
itself such as “pinched together” peaks/troughs and discontinuities. An illustrative
analogy is the crumpling of a two-dimensional piece of paper resulting in complex
three-dimensional structure. The pores in the activated carbon are the interstices in
this three-dimensional topology, as depicted in figure 2.3.
20 Literature Review
2.3.2 Activation Process
While material carbonised as described above has porous structure, this is insufficient
for commercial applications. In order to make activated carbon, the carbon undergoes
a process (eponymously, “activation”), which creates further porosity, widens existing
porosity, opens access to previously inaccessible pores, and modifies the surface of the
carbon. This is achieved by either thermal or chemical means, or a combination of the
two.
2.3.2.1 Thermal Activation
“Thermal” activation involves gasifying the carbon with the oxidising gases, steam, or
CO2, at high temperatures (>900°C), according to the following equations:
C+ CO2 = 2CO (2.1)
C+H2O = CO+H2 (2.2)
Oxygen gas is not used as its combination with carbon is highly exothermic, diffi-
cult to control, and results in consumption of the carbon (or burning) rather than
penetration into, and enhancement of, existing pores. In this way, carbon atoms
are selectively removed from the molecular surface. The carbonisation pyrolysis
process can create tars and other decomposition products that fill or partially block
pores; this more disorganised carbon is the first to react with the oxidising gases,
creating new accessible porosity by “clearing” the blocked pores [36]. These reactions
also widen existing rudimentary porosity of the char, abstracting atoms from the
carbon structure, and creating defects in the lattice such as free valencies, which
contribute to activated carbon’s properties. Furthermore, the reaction of the oxygen-
containing gases creates surface oxygen complexes, which themselves have reactivity
and contribute to adsorption [52].
2.3.2.2 Chemical Activation
Activation can also proceed via chemical means, where the carbonisation and activation
occur simultaneously. This involves impregnating the source material with a chemical
agent, heating in an inert atmosphere, and then washing any remaining chemical
out. The most commonly used agents for chemical activation are ZnCl2, H3PO4,
and KOH. ZnCl2 (which at high concentrations exhibits Bronsted acidity) and H3PO4
2.3 Activated Carbon 21
activation proceed via acid-catalysed dehydration and depolymerisation, promoting
bond cleavage and the formation of cross-links (with cyclisation and condensation
reactions) [72–74]. KOH activation occurs via a redox reaction where carbon is oxidised
to CO or CO2, with the reduction of the potassium to a metal ion that intercalates
into the carbon structure and expands the lattice; destructive removal of this ion
occurs when the carbon is progressively heated, creating new pores by disrupting the
graphene structure as it exits [73, 75]. The different chemical routes produce activated
carbon with different porous profiles, but the end result in all cases is a highly porous
active carbon.
2.3.3 Applications of Activated Carbon
Activated carbon has innumerable applications, though the bulk of these revolve around
its superlative ability to physically adsorb chemical species as a result of its highly
developed porous structure.
2.3.3.1 Commercial Applications of Activated Carbon
Water treatment is the single largest use of activated carbon worldwide, with the
primary focus being the removal of pollutant organic compounds, either to create
potable water or to remove toxic and harmful contaminants before discharge to the
environment. Specific applications include the removal of volatile organic compounds
[76–78], pharmaceuticals [79–81], natural organic matter [82–84], and heavy metals
[85–87]. Activated carbon is also used significantly in the food and beverage industries
to remove colour or odour from products such as fruit juices [88], sugar [89, 90],
and vegetable oil [91]. Activated carbon’s excellent adsorptive ability is also used
in chemical and pharmaceutical production, especially in fermentation processes
where antibiotics, vitamins, and steroids are adsorbed in the activated carbon and
then recovered by solvent extraction followed by distillation [52, 92, 93]. The two
largest gas-phase applications of activated carbon are solvent vapour recovery [94]
and automotive fuel evaporative loss systems [52].
2.3.3.2 Activated Carbon as a Catalyst Support
Activated carbon is widely used as a catalyst substrate, supporting the catalyst and
allowing it to be spread over the activated carbon’s large surface area, to increase
the number of catalytically active atoms exposed to substrate. Activated carbon has
22 Literature Review
particular utility in this role as it is resistant to acidic and basic media, is stable up to
very high temperatures, and allows for the easy recovery of the active phase by simply
burning the carbon off [95]. The catalyst is introduced by bringing a solution containing
the active phase (or a suitable precursor) into contact with the activated carbon, where
it adsorbs onto the surface. The mixture is dried and heat treated (sintered) to stabilise
the catalyst [96]. The role of surface oxygen complexes is also important, with catalyst
dispersion and sintering increasing with the number of oxygen surface groups [97, 98].
The number of reactions mediated by activated carbon supported catalysts is
extremely large. Auer et al. [99] list a “selected” 40 classes of reaction catalysed by Pd,
Pt, Ir, Ru, and Rh supported by activated carbon. Examples dealing with the production
of hydrocarbons include the Fischer-Tropsch synthesis of synthetic fuels with Mo, Co,
and Fe based catalysts [100–102]; the pyrolysis of vegetable oils into a diesel-like liquid
using activated carbon supported Pd [103]; and the conversion of palm oil to biodiesel
using activated carbon supported KOH [104].
2.3.4 Activated Carbon as a Catalyst
In recent times there has been growing appreciation of activated carbon as a useful
catalyst in its own right, and it is currently used commercially in the production of
carbonyl chloride from carbon monoxide and chlorine [36]. While work using activated
carbon alone as a catalyst is relatively sparse in the literature, there is a slowly growing
body of such reactions. Trimm [37] presented a list of reactions known to be catalysed
by activated carbon, which is summarised in table 2.1.
One of the areas in which concerted activated carbon research has taken place
is in the oxidative removal of harmful contaminants from water and gas streams.
In these applications the high porosity of activated carbon facilitates capture of
Reactions Involving Examples
Hydrogen RX+H2 → RH+HX (X = Cl, Br)
HCOOH→ CO2 +H2
CH3CHOHCH3 → CH3COCH3 +H2
Oxygen SO2 + 12O2 → SO3
NO+ 12O2 → NO2
2H2S + O2 → S2 + 2H2O
Halogens CO+Cl2 → COCl2
C2H4 + 5Cl2 → C2Cl6 + 4HCl
SO2 +Cl2 → SO2Cl2
C6H5CH3 +Cl2 → C6H5CH2Cl + HCl
Table 2.1: Reactions catalysed by activated carbon. Reproduced from [36, 37].
2.3 Activated Carbon 23
even low concentrations of substrate, which is then catalytically oxidised with high
efficiency, resulting in a purified output stream. This process has been successfully
demonstrated in the oxidative removal of H2S, SO2, and NOx [105–107], with these
reactions proceeding along the pathways outlined in table 2.1.
The ability of activated carbon to catalyse the oxidative dehydrogenation of sub-
stances has been studied in order to produce various chemical feedstock, such
as the conversion of isobutane to isobutene [108], the highly selective oxidative
dehydrogenation of alkylbenzenes to produce styrene (the monomer of polystyrene)
and α-methylstyrene (the precursor to a number of plasticizers and resins) [38, 109–
113], as well as the dehydrogenation and dehydration of various alcohols [114–116].
2.3.4.1 Catalytic Hydrogen Production using Activated Carbon
Another area that has seen the application of activated carbon as a catalyst is the
production of hydrogen from methane, according to the reaction:
CH4 → C+ 2H2 (2.3)
This process has the potential to produce hydrogen with significantly reduced CO2
emissions, and is a substantially simpler operation than many conventional hydrogen
production processes such as the steam reforming of methane. Muradov et al. have
successfully demonstrated this process, espousing the virtues of activated carbon as a
catalyst because of its high availability and durability, low cost, and advantages over
metal catalysts such as high temperature resistance, and tolerance to sulphur and other
harmful impurities that often arise from natural hydrocarbon feedstock [39, 117–121].
While activated carbon has higher activity than other comparable catalysts for this
purpose, it has been noted that activated carbon does deactivate over time as a result
of carbon deposition from the hydrocarbon decomposition [122].
In a related application, Petkovic et al. [123] have used activated carbon to decom-
pose hydrogen iodide to produce hydrogen, a key reaction in a wider thermochemical
scheme to split water into hydrogen more efficiently than through pure thermal means.
24 Literature Review
2.4 Microwave Heating
The total charge of a system in an electric field is given by:
D = εE (2.4)
where D is the electric charge density vector (C/m2), E is the externally applied electric
field (V/m), and ε is the absolute permittivity (F/m) — the measure of resistance that is
encountered when forming an electric field in a medium. The permittivity of a material
is usually given relative to that of free space:
ε = εrε0 (2.5)
where εr is the relative permittivity, and ε0 is the permittivity of free space (8.86 x 10-12
F/m). A material’s polarisation does not respond instantaneously to an applied field,
and is dependent on (angular) frequency (ω); it is then convenient to treat permittivity
(either relative or absolute) as a complex function:
ε(ω) = ε′(ω)− iε′′(ω) (2.6)
where ε’ is the real part of the permittivity, known as the dielectric constant, and is
related to the ability of a material to store an electric field’s energy (e.g., a microwave
field inducing dipoles), and ε” is the imaginary part of the permittivity, known as the
dielectric loss factor, which is governed by the lag between the application of an
electric field and a material’s polarisation, and represents the dissipation of energy
(as heat) within the medium. This energy loss appears as an attenuation of the applied
field and is often measured as the ratio of the dielectric loss factor to the dielectric
constant, or the angle between the imaginary axis and the total current density vector
(i.e., the sum of the real and imaginary components of the permittivity), the dielectric
loss tangent or dissipation factor:
tanδ =
ε′′
ε′
(2.7)
Both the dielectric loss factor and the dielectric constant can vary with the frequency
of the applied field and the temperature of the material, and a material with a large
dielectric loss tangent (i.e., a large dielectric loss factor and moderate dielectric
constant) will be a good microwave receptor, receiving and converting this energy to
heat with high efficiency [124–126].
Briefly, there are three mechanisms by which microwave heating occurs. Dipole reori-
2.4 Microwave Heating 25
entation occurs when the chemical bonds of dipoles are realigned by the fluctuating
electromagnetic field. This movement creates friction between the rotating molecules
and causes heat to be generated throughout the volume of the material. The dipoles
themselves can be innate to the molecule, arising from asymmetric charge distribution,
or induced by the microwave field, displacing electrons or nuclei from their equilibrium
position. Similar interactions with the electromagnetic field can also occur as a result
of interfacial, or Maxwell-Wagner polarisation, where charge builds up in contact areas
or interface layers in heterogeneous materials as a result of different conductivities
or dielectric properties. Finally, conductive heating occurs when an electrically
conductive material is exposed to an electromagnetic field, inducing charged particles
(e.g. electrons, ions) to move. Heating occurs as a result of the material’s innate
resistance to this flow of current.
2.4.1 Dielectric Properties of Materials and Influence on
Microwave Receptivity
Plastics are poor microwave receptors — the dielectric loss tangent of polyethylene is
0.0007 [127] (all such values in this text are listed at room temperature and with a
microwave field at 2.45 GHz); by way of comparison, distilled water — an excellent
microwave absorber — has a dielectric loss tangent of 0.118 [128]. As long non-polar
molecules, plastics polarise and reorientate poorly; being homogeneous, Maxwell-
Wagner polarisation does not occur with pure plastics, and with no free electrons
nor does conductive heating. Clearly, plastics are very poor microwave receptors,
and as such are unable to be directly targeted safely by microwave heating. Instead,
when heating plastics (or indeed any microwave transparent material) with microwave
radiation an intermediate receptor is used and the energy is transferred to the plastics
via short-range conduction.
2.4.1.1 Carbonaceous Materials as Microwave Receptors
The microwave receptivity of carbonaceous materials varies depending on their chem-
ical composition. Diamond, for example is almost entirely microwave transparent
owing to its rigid lattice-like structure deriving from its sp3 bonding, and lack of
free electrons. Graphite, on the other hand, with its sp2 bonding and delocalised
electrons, heats well when exposed to a microwave field, primarily through conductive
heating. Carbonaceous materials with a high level of structural irregularity such as
26 Literature Review
coal (assuming no moisture content) are relatively poor microwave receptors (with a
dielectric loss tangent of 0.02-0.08 [128]), owing to the lack of delocalised electrons;
however, as the structural order increases so too does the dielectric loss tangent and
microwave receptivity. Coke and activated carbon, which aside from the potential
catalytic and reducing abilities that make them well suited to pyrolysis, have high
dielectric loss factors of 0.35–0.83 and 0.22–2.95 respectively [128, 129], making
them good microwave receptors. Furthermore, these materials are generally relatively
inexpensive, and are easily available in different textures, sizes, forms, etc., all of which
contributes to their selection for use in this work.
2.4.1.2 Power and Penetration Depth
The power transmitted to an object in a microwave field can be calculated from the
Maxwell equations: assuming a uniform electric field, E, the power, P, absorbed per
unit volume is given as:
P = 2pifε0ε
′′E2 (2.8)
i.e., absorbed power increases with the square of the strength of the electric field.
However, as energy is absorbed by a material the electric field strength decreases,
making the above equation only valid for thin materials, if they are microwave
absorbent. The penetration depth of a microwave field is usually defined as the
distance below the surface where the absorbed power is 1/e (i.e., approximately 37%)
of that of the power at the surface of the material [125, 130]. In many cases when
highly microwave-absorbent materials are placed in a “normal” strength microwave
field (such as the carbonaceous microwave receptors described in this chapter) it is
unlikely that the microwave field would penetrate throughout the entire material, a
further reason that stirring is often in employed in these cases.
2.4.2 Microwave Heating Modes
2.4.2.1 Single-Mode Microwave Heating
Single mode microwave heating occurs in cavities where the incident and reflected
waves give rise to a single mode standing wave pattern. For simple designs this form-
ation can be solved analytically, giving a predictable electromagnetic field distribution
in the cavity. This necessitates a microwave source with very little variation in output
frequency, and with one resonant mode the size of these applicators is approximately
the order of one wavelength, with one “hot spot” where the microwave field strength is
2.4 Microwave Heating 27
very high [130]. The small size and large power intensities achieved with this form of
heating lead to relatively niche applications such as joining ceramics [131], and curing
thermosetting resins [132], though these properties also make it difficult to operate
these reactors on anything other than small scale.
2.4.2.2 Multi-Mode Microwave Heating
In multi-mode heating, microwaves are channelled into a cavity with dimensions
substantially larger than that of the microwave’s wavelength. This results in a large
number of standing wave modes that can exist at or near to the operating frequency.
To minimise the formation of “hot” and “cold” spots (high and low field strength) in the
cavity as a result of constructive and destructive interference occurring it is desirable
to excite as many of these modes as possible. In spite of this, heating uniformity can
often be difficult to achieve in multi-mode ovens and this often promoted through the
use of metallic mode stirrers or by rotating the product within the field [130]. The
ability of multi-mode heating to cope with larger objects and batch operations make it
by far the most common form in industrial applications.
2.4.3 Microwave vs. Thermal Effects
Microwave heating has been employed to facilitate a wide range of reactions (in a
rather comprehensive review of microwave-assisted organic synthesis, Lidström et al.
[133] individually listed over 600 separate reactions enhanced by microwave heating).
Studies of microwave-assisted reactions have noted them proceeding faster (by 2–3
orders of magnitude), more cleanly, and with higher yields than their conventionally
heated counterparts [134–136]. However, the source of these differences is the
subject of some controversy, with some researchers positing microwave-specific effects
over and above those that can be accounted for by heating and temperature profile
differences alone. For example, there is some compelling evidence that microwave
fields enhance sintering reactions by providing a voltage-induced driving force that in-
creases grain boundary diffusion [137]; Marand et al. [132] have advanced explanations
accounting for the microwave-enhancement of organic synthesis reactions in which
preferential orientation of (dipole-containing) molecules are induced by the microwave
field, or in which polar reactive groups are selectively activated (as opposed to heating
the entire molecule as would be the case in externally derived conductive heating); de la
Hoz et al. [138] provide a good review of various research groups’ evidence in support
of microwave fields providing some sort of sub-molecular selective energy transfer.
28 Literature Review
In some instances, though, the case for microwave-specific effects has been over-
stated, with a critical re-examination of various reactions where previous authors
had claimed microwave-specific influences showing no evidence for the existence
of non-thermal microwave effects [139, 140]. In this re-examination, the original
discrepancies were attributed to inadequate stirring resulting in inhomogeneous
temperature distribution, and/or less accurate temperature measurements such as
those arising from infrared sensors, which do not provide accurate readings in
the microwave heating context (by contrast, the re-examination experiments were
conducted in a reactor that used multiple highly accurate fibre-optic temperature
probes). In the case of liquids and gases, even highly targeted microwave energy is
unlikely to cause purely thermal microwave-specific effects, as redistribution of kinetic
energy between molecules occurs more rapidly than the conversion of electromagnetic
energy into kinetic energy [141]. However, it should be noted that this fact does not
preclude non-thermal microwave-mediated interaction.
While there is wide acceptance of the ability of microwave heating to selectively
heat substances on the macroscopic (and even molecular) scale, whether this extends
down to direct activation of functional groups or other non-thermal effects remains to
be conclusively settled. Independent of whether this occurs, it is clear that microwave
heating does result in different thermal heating profiles to conventional techniques
(e.g., fast heating rates, volume heating, high power density, material selectivity, even
heating, and good heating control) and this is of demonstrable significant value in and
of itself.
2.4.3.1 Microplasmas
An instance of an observable microwave-specific phenomenon directly applicable to
this study is the formation of microplasmas [29, 142–144]. In carbon materials
with graphite-like structural elements, where there are a large number of relatively
mobile delocalised electrons, the microwave field may induce charge imbalances that
equalise across particle boundaries in an arc that conductively ionises the intervening
atmosphere. These microscopic plasmas present high temperatures and extreme (if
localised) conditions that have the potential to influence the course of the bulk reaction.
2.5 Microwave-Assisted Pyrolysis 29
2.5 Microwave-Assisted Pyrolysis
Microwave-assisted pyrolysis has been applied to a wide variety of inputs, resulting in
products that show a range of advantages over the source material such as conversion
or separation into less hazardous forms, significant volume reduction of waste, and the
production of useful fuels or chemical feedstock. An illustrative set of examples of these
processes is presented below.
2.5.1 Processing of Problematic Waste
2.5.1.1 Scrap Tyres
As a complex composite material that is expendable, produced in large quantities, and
potentially environmentally hazardous, automotive tyres present a challenging disposal
problem. A number of research groups have examined the possibility of recycling scrap
tyres using microwave-assisted pyrolysis, recovering a gas and oil fuel product, steel,
and high value activated carbon or carbon black [145–148]. Furthermore, the absence
of oxygen in the pyrolysis process means NOx, SOx and other harmful organic emissions
are significantly reduced compared with disposal by incineration. The carbon black
content in tyre rubber makes this material a sufficiently good microwave receptor that
an intermediate bed need not be used.
2.5.1.2 Used Automotive Engine Oil
Lam et al. [149–152] have used microwave pyrolysis to recycle used automotive
engine oil, a hazardous waste containing heavy metals and polyaromatic hydrocarbons.
Pyrolysis of this material using a coke bed at temperatures between 400–700°C resulted
in a high recovery of the calorific and chemical value of the oil, with a condensed
product consisting of aliphatics and aromatics, and a non-condensable gaseous fraction
containing light hydrocarbons and some hydrogen. The microwave pyrolysis also acted
as a separations process, with the majority of the heavy metals and PAHs becoming
embedded in a char fraction that was separable from the reactor bed.
2.5.1.3 Sewage Sludge
Sewage sludge is another hazardous waste that has been processed using microwave
pyrolysis in order to generate syngas (H2 + CO). In these studies, led by Menéndez et al.
at the Instituto Nacional de Carbon in Spain [21, 153–155], the sludge was initially
30 Literature Review
pyrolysed with a fresh coke bed, but the char that was produced as a result of the
process was also demonstrated to be an adequate microwave receptor, and was used in
subsequent runs. The water content of the sludge was used to gasify the material, with
a final operating temperature of 1040°C and typical yields of 62% aqueous, 25% gas,
10% char, and 3% oil. The primary components of the gas were H2, CO and CH4 with
percentages of 22–43%, 23–30% and 2–7%, respectively. The use of microwave-assisted
pyrolysis resulted in a larger gas fraction, and lower polycyclic aromatic hydrocarbon
content compared with conventionally heated sludge [156, 157].
2.5.2 Microwave-Assisted Pyrolysis of Biomass Waste
Aside from the sewage sludge already mentioned, a wide variety of novel biomass has
been the subject of microwave-assisted pyrolysis research, including wood [19, 158],
coffee hulls [12], micro-algae [159], rice straw [160], waste tea [161], and wheat straw
[162]. These experiments used non-catalytic carbon beds when processing materials
that were poor receptors of microwaves, though others (such as the pyrolysis of wood)
heated the material directly. A variety of gas, liquid and solid phase products resulted,
these being determined by the characteristics of the original feedstock and the
processing conditions. Potentially useful gaseous and liquid products were produced,
though in many cases these would likely require further separation and/or refining
before end use. The bulk heating characteristics of microwave-assisted pyrolysis allows
the processing of materials without energy intensive and costly grinding/shredding
(e.g., entire large wood blocks were processed by Miura et al. [158]), limited only by the
size of the reactor and the ability of the microwave field to penetrate the material. This
fact, combined with the more rapid reaction rate that was observed in the cases where
microwave pyrolysis was compared with conventional electric heating, would result
in a large overall reduction in electricity consumption compared with conventional
processing.
2.5.3 Upgrading Naturally Occurring Hydrocarbon Sources
Microwave-assisted pyrolysis has also been applied to a number of naturally occurring
hydrocarbon materials in order to produce fuels by extracting volatile components
and cracking long-chain hydrocarbons. Oil shale, for example, has been subjected to
microwave pyrolysis at 700°C, yielding around 6 wt.% of a highly aromatic oil and
10 wt.% gaseous products [163]. Similar yields were obtained using conventional
2.5 Microwave-Assisted Pyrolysis 31
heating but the composition of the products differed, with more cracking occurring
in the microwave-assisted pyrolysis case, and with fewer sulfur, nitrogen, and polar
compounds in the resulting products.
Coal has also been processed by microwave-assisted pyrolysis at temperatures up
to 1300°C in order to produce condensible tars for use as a fuel [164, 165]. While coal
can be heated directly using microwave radiation, it is a relatively poor microwave-
absorbent, and in this study a variety of different microwave intermediates were used
to increase the pyrolysis reaction rate. Yields of condensible tars varied depending on
the intermediate microwave receptor used: with coke the tar yield was 20 wt.%, using
Fe3O4 resulted in 27 wt.%, and a maximum of 49 wt.% was seen with the use of CuO.
The tars were primarily aliphatic (65–74% of the oil), with the remainder consisting of
aromatic species.
2.5.4 Microwave-Assisted Pyrolysis with an Activated Carbon
Bed
There have been a limited number of studies specifically using activated carbon in
conjunction with microwave-assisted pyrolysis, and these are briefly discussed as
follows. Ma et al. [166] have used microwave-assisted pyrolysis over activated carbon
in order to remove NO from simulated combustion flue gas through conversion to N2.
The use of microwave heating in this context facilitated a more rapid conversion, and
preserved the porous structure of the activated carbon more effectively, compared with
conventional heating.
Research carried out at the Instituto Nacional de Carbon has employed microwave-
assisted pyrolysis to crack glycerol into syngas at 800°C [22, 167]. Using an activated
carbon bed produced slightly higher overall gaseous yields, a greater syngas yield, and
a CO/H2 ratio closer to one, when compared with the same procedure using an inert
quartz blank.
The same group has used activated carbon as a catalyst in the microwave-assisted
decomposition of methane to hydrogen [142, 168]. With an inert control bed, no
decomposition of methane occurred at 800°C, though this occurred readily over the
activated carbon bed. Furthermore, even when methane decomposition did occur in
the non-catalytic case at higher temperatures, the selectivity for H2 production was
improved by nearly 100% through the use of activated carbon. The combination of
activated carbon and microwave heating resulted in greater yields than pyrolysis with
other carbon materials, and with conventional heating using activated carbon; this
32 Literature Review
was attributed to the formation of microwave-induced microplasmas in the bed. In
longer experiments of two hours’ duration some deactivation of the activated carbon
was noticed, which was attributed to carbon deposits entering the pores and reducing
the surface area.
Lignocellulosic biomass has also been subjected to microwave-assisted pyrolysis
using activated carbon, with Douglas fir pine producing an aromatic oil at 25–35 wt.%
yield. The use of activated carbon resulted in an oil with a higher content of esters and
phenolic species compared with the non-catalysed case [169].
2.6 Pyrolysis of Plastics
The pyrolysis of plastics, with the aim of recovering valuable chemical compounds,
has been carried out by a number of researchers using a wide variety of source
material, though almost all of this work has been conducted with conventional heating
techniques; a review of these studies is presented below. The percentages given in this
section are all weight percentage of input plastic, unless specified otherwise.
2.6.1 Pyrolysis of Polyethylene
Research into the pyrolysis of plastics has been carried out at the University of
Hamburg, led by Walter Kaminsky, since the 1970s [170]. This research has seen
the development of a series of fluidised bed reactors (FBRs), in increasing sizes from
laboratory plant (100 g/h), pilot plant (30 kg/h) through to prototype plant (150 kg/h).
This configuration, referred to as the Hamburg process, recirculates a portion of the
produced pyrolysis gas to fluidise the bed. Applying this process to polyethylene at an
operating temperature of 830°C resulted in gaseous and condensed pyrolysis products,
the four most prevalent species being, in order: 27.6% ethene, 23.7% methane, 16.2%
benzene, and 4.0% toluene [171]. A separate study at 740°C saw the products comprise
55.8% gas, 42.4% oil and 1.8% residue (char), with the equivalent prevalent species:
23.8% methane, 20.0% ethene, 19.2% benzene, 3.9% toluene.
Within the U.K., the Williams group at the University of Leeds has led research into
the pyrolysis of plastics. When pyrolysing HDPE in a fixed bed reactor, increasing from
room temperature to 700°C at a rate of 25°C/min, the recovered pyrolysis products
included 17% gases, 80% oil, and no char. FTIR spectra of the pyrolysis oil indicated
it was primarily aliphatic [172]. Pyrolysing LDPE in a fluidised bed at a range of
temperatures from 500–700°C resulted in steadily increasing gas yields with increasing
2.6 Pyrolysis of Plastics 33
temperature from 10.8% to 71.4%, while oils and waxes fell from 89.2% to 28.6% over
the same range. The gases were short-chain linear aliphatics, with methane, ethene,
propene, and butene increasing the most as the reaction temperature increased. The
oils and waxes were aliphatic at 500°C and 550°C, with aromatics first detected at
600°C and increasing in concentration thereafter [173].
Conesa et al. [174] have used an FBR to pyrolyse HDPE at temperatures between
500–900°C. As reaction temperature increased so did cracking and yield of gases,
which reached a maximum of 94.5% at 800°C, where the main components of the
pyrolysis products were ethene (37%), benzene (24%), methane (20%), and propene
(4.7%). At 900°C the total yield declined slightly, a fact attributed to the cracking of
input sample to soot. Comparing the FBR results to that obtained using a Pyroprobe
1000, in which secondary reactions and residence time were minimal owing to the
small physical area that was heated, showed significantly less cracking than when the
FBR was used at equivalent temperatures, and a change in composition with minimal
aromatics present (the maximum observed yield of benzene was 2.3 wt.%), highlighting
the importance of these factors in both overall cracking and aromatic formation.
Milne et al. [175] pyrolysed LDPE in an internally circulating FBR at 800°C.
Changing the residence time from 400 ms to 600 ms resulted in a dramatic change in
product composition, with increases in methane (10.9% vs. 23.3%) and ethene (18.7%
vs. 32.3%), and decreases in heavier C4 species (42.6% vs. 13.1%). Minimal char
(~2%) and condensate (5–8%) were produced, though this was not analysed further.
Aguado et al. [176] used a conical spouted bed reactor to separately pyrolyse both
LDPE and HDPE between 450–600°C, and found little difference between the resulting
pyrolysis products. The yield of pyrolysis waxes decreased almost linearly from 80%
to 49% over this range, with the gas yield making up the remaining fraction. Both of
these components comprised linear aliphatic species.
Finally, Mastral et al. at the University of Zaragoza in Spain have conducted research
into the pyrolysis of HDPE, using an FBR at temperatures between 640–850°C. At
640°C the obtained yields were 78.9 wt.% waxes/oils and 11.4 wt.% gases, with the
waxes consisting of aliphatic hydrocarbons. Cracking and gas yields increased with
temperatures to a maximum at 780°C (86.4% gases, 9.6% oils). At this temperature
the most common gases identified were methane (14.2%) and ethene (41.9%). The
quantity of aromatics increased with temperature, ranging from 1.5% at 640°C to
29.7% at 850°C. No char was produced in these experiments [11, 177]. A separate
series of experiments [178, 179] demonstrated the importance of temperature and
34 Literature Review
residence time on the product distribution using two free-fall reactors in series.
This allowed primary and secondary reactions to be studied separately, with primary
pyrolysis reactions occurring in the first reactor at 600°C, and secondary cracking
and recombination taking place in the second reactor at higher temperatures. With
both reactors operating at 600°C aliphatic products were produced and aromatic
species were insubstantial (<0.2%); however, aromatic species increased dramatically
to around 50% or greater when the second reactor was raised to a high temperature
(1000°C) or, at a lower temperature but with a longer residence time (800°C with
the residence time increased from 1.8 s to 5.8 s). Greater overall cracking, including
greater gas yields, was also observed with a longer residence time at a given
temperature.
2.6.1.1 Microwave-Assisted Pyrolysis of HDPE
With the exception of the Hamburg process, which can also be gas powered in some
configurations, all the work described above used conventional electric furnaces as
heat sources to supply energy for the pyrolysis. The use of microwave heating for
the pyrolysis of plastic has been pioneered by Ludlow-Palafox and Chase [29, 46]
at the University of Cambridge, and has been so far limited to their studies. Their
investigations used a 5 kW modified catering oven, within which HDPE was pyrolysed
in a quartz stirred-bed reactor. Yields of waxes and gases were found to be similar at
500°C and 600°C, at around 80% and 20% respectively. This similarity was attributed to
the increased cracking rate at 600°C being counterbalanced by a decrease in residence
time as a result of pyrolysis gas being more rapidly forced out of the reactor through
volume expansion at the higher temperature. Both the collected gases and waxes were
primarily linear aliphatics, with the waxes comprised of linear alkanes, alkenes and
alkadienes, with minimal aromatic content (<2%). The microwave-assisted pyrolysis
process was also successfully applied to aluminium-plastic laminate (“toothpaste tube”)
with successful recovery of both the hydrocarbon and aluminium fractions, an ability
attributed to the “gentle” microwave heating mode. The potential of these processes is
such that a successful spin-off company has been formed to develop them commercially,
securing funding for a pilot plant that is currently operational [180].
2.6.1.2 HDPE Pyrolysis Overview
While the preceding review may seem something of a barrage of data, it does serve to
illustrate the variety of approaches that have been taken in researching the pyrolysis of
2.6 Pyrolysis of Plastics 35
CH3
n
C
H
HH
C
n
C C
ClH
H H n
Polypropylene Poly(vinyl chloride)Polystyrene Poly(methyl methacrylate)
OO
n
Figure 2.4: Chemical structures of some plastics.
HDPE, and the sensitivity of the pyrolysis process to a range of factors including reactor
type, residence time, operating temperature, type of reactor bed, and other process
variables. Nevertheless, there are certain overarching trends that can be drawn from
the above survey:
• The pyrolysis of HDPE results in decomposed shorter chain hydrocarbons, com-
prised of gases, and waxes and oils, with small amounts of char sometimes forming
at high temperatures.
• Increasing reaction temperature generally increases overall cracking, resulting in
a greater yield of gases and fewer oil/wax products.
• The produced gases are comprised of C1–C5 aliphatics, with methane and olefins
proportionally increasing the most with increased reaction temperature.
• The oils and waxes are comprised of linear aliphatic species at lower reaction
temperatures, though above 600°C aromatic species are formed and can comprise
a significant portion of the produced products.
• The residence time of products in the reactor has a similar effect to that of the
reaction temperature, with increases in either resulting in more cracking and
greater gas and aromatic yield.
The relevant data from the studies presented above are summarised in tabular form
in §4.8.1, where they are compared in detail with those obtained in the course of
this study, accompanied by explanations accounting for any unexpected differences
between the two.
2.6.2 Pyrolysis of Other Plastics
While this work uses HDPE as a model plastic, it is certainly possible to pyrolyse other
sorts of plastic, and work has been carried out to this end. Illustrative examples of
pyrolytic recycling of some of the more common plastics are given below; the structures
of these polymers is given in figure 2.4.
36 Literature Review
2.6.2.1 Polypropylene
The pyrolysis of polypropylene (PP) produces aliphatic products in the same manner
as HDPE; however, the branching inherent to the polymer structure results in highly
branched pyrolysis products. Bockhorn et al. [181] pyrolysed PP in a crucible at
temperatures between 410–460°C, resulting in a diverse variety of aliphatic pyrolysis
products, comprising 7.6% alkadienes, 7.6% alkanes, and 84.8% alkenes, most of which
were highly branched.
2.6.2.2 Polystyrene
The pyrolysis of polystyrene (PS) is able to recover a significant fraction of the original
styrene monomer. In this case the styrene monomer is a relatively stable unit, and
decomposition proceeds by scission of the intervening C-C bond, which has a lower
bond dissociation energy than either the benzene ring or the benzene-carbon bond.
Aguado et al. [182] pyrolysed PS in a conical spouted bed reactor (CSBR) with a sand
bed, and were able to recover a maximum yield of styrene monomer of 64.5 wt.% at
475°C. The remaining collected products were mostly aromatics, except for a small
fraction of gases that had been cracked below C6, necessarily disrupting the benzene
ring. This compares with the work of Scott et al. [183], who used a fluidised bed, and
gained a maximum monomer yield of approximately 76 wt.% at 532°C, with a total
aromatic yield of 88 wt.%, with the remainder comprising non-condensable gases.
2.6.2.3 Poly(vinyl chloride)
The chlorine in poly(vinyl chloride) (PVC) can cause significant difficulty when pyro-
lysing PVC. The C-Cl bond dissociation energy is lower than others in structure; as
a result, degradation of the polymer proceeds via chain-stripping of the Cl [13] in
a dehydrochlorination reaction that results in hydrochloric acid, which can cause
corrosion and damage to apparatus [183].
Williams and Williams [172] pyrolysed PVC in a fixed bed increasing from room
temperature to 700°C at a rate of 25°C/min. HCl gas from the dehydrochlorination
reaction was produced from 150°C, and was the primary overall pyrolysis product
at 52.9 wt.%. After the chlorine has been removed from the polymer, degradation
proceeds as standard non-chlorinated vinyl, with the remainder of the pyrolysis
products in this study (2.5 wt.% gas, 31.7 wt.% oil) comprising aliphatic species, with a
2.6 Pyrolysis of Plastics 37
small quantity of aromatics. 13.8% of the input mass resulted in the formation of char
due to interaction with the HCl.
The differing temperatures at which dehydrochlorination and polymer backbone
scission occur allow the pyrolysis of PVC to be performed in a stepwise sequence
with the HCl separated first before the remaining hydrocarbon pyrolysis products are
formed. A multi-stage circulated-sphere apparatus used by Bockhorn et al. [184] was
able to strip 99.6% of the Cl from the PVC at 330°C, with pyrolysis of the remaining
fraction occurring at 440°C and resulting in aliphatic species.
2.6.2.4 Poly(methyl methacrylate)
Poly(methyl methacrylate) (PMMA) is an example of a plastic from which especially
high monomer recovery can be achieved using pyrolytic recycling. In this case the
ester monomer forms a particularly stable unit, and decomposition occurs via end chain
scission back into the original monomer [13].
Using the fluidised bed Hamburg process at 450°C, Kaminsky and Franck [185] were
able to recover 97 wt.% of input PMMA as methylmethacrylate monomer; however,
this process was highly temperature dependent, declining to 55% at 590°C, with
the remaining fraction consisting of CO/CO2 and short-chain gaseous hydrocarbons,
indicating decomposition of the ester group had occurred.
2.6.2.5 Effect of Monomer on Pyrolysis Products
The overarching observation from the above examples is that the composition of the
pyrolysis products is highly dependent on the chemical structure and bond energies
of the components of the polymer. While monomer recovery varies significantly from
plastic to plastic, this is not to say that pyrolysis products in forms other than the
monomer are not valuable; indeed, they are able to be applied to a diversity of other
applications as chemical feedstock or fuels.
2.6.3 Catalytic Pyrolysis of HDPE
Catalysts other than activated carbon have also been used to pyrolyse plastics, with
most of these deriving from the petrochemical refining industry. Most of these
substances are acid-catalysts, with surface Lewis and Bronsted acid sites, which
abstract hydrides and donate protons respectively, and lead to different reaction
pathways than pure thermal cracking. The mechanism for the acid-cracking of
38 Literature Review
hydrocarbons is well developed and described in detail in [13]. Given the large number
of publications using these catalysts, rather than present a comprehensive description
of individual research, this section synthesises trends across the literature surveyed for
the various catalyst types.
2.6.3.1 Zeolites
Zeolites are microporous aluminosilicate minerals and are probably the most studied
catalysts in the cracking of hydrocarbons. They exhibit high stability and high acid
strength, and as a result of their microporous structure exhibit selectivity for smaller
molecules (the porous structure acting as a molecular sieve with larger molecules
physically unable to diffuse into the pores) [186]. Multiple types of zeolite exist, and
these differ in chemical composition and in the structure of the underlying framework,
with different resulting silicon/aluminium ratios (Si/Al). Two common types of zeolite
are Y-zeolite (which has a pore formed by a twelve membered ring and a Si/Al ratio
from 1.5–3.0), and ZSM-5 (which has a pore formed by a ten membered ring and a
Si/Al ratio >10); both are commonly used as cracking catalysts in the petrochemical
industry. The Si/Al ratio determines the surface acidity of the catalyst by altering the
surface concentration of aluminium, which is the source of the acid sites that underlie
catalytic activity [187].
The cracking of HDPE with zeolites is able to occur at significantly lower reaction
temperatures than non-catalysed pyrolysis, with zeolite-catalysed reactions exhibiting
more rapid reaction rates and greater cracking, resulting in lighter products and
larger gaseous yields; these increase with increasing temperature [188]. Other
characteristics of zeolite-catalysed cracking of HDPE include high yields of aromatics
and olefins in the pyrolysis products, as well as isomerisation and branched products
[188–190]. Aromatic species are considerably more prevalent when using Y-zeolite
as opposed to ZSM-5, attributed to Y-zeolite’s lower Si/Al ratio [191], but also likely
resulting from the greater selectivity of ZSM-5 for smaller molecules (resulting from
its smaller pore size), which preferably generates C3–C5 species as opposed to C3–C8
with Y-zeolite [192]. Ultimately the pyrolytic products produced when using zeolite
are reflective of its chemical and physical make-up (i.e., surface acidity and pore size),
which can be tuned towards specific desired outcomes [193].
2.6 Pyrolysis of Plastics 39
2.6.3.2 Aluminosilicates
Aluminosilicates are distinguished from zeolites by their lack of microporous structure,
but they do exhibit a wide distribution of larger meso- and macropores. As a result they
have lower surface area and lower strength acid sites than zeolite, with correspondingly
less catalytic activity [186]. The larger pores allow larger molecules to diffuse into
and out of pores, giving these catalysts less selectivity than zeolites, which has been
demonstrated in the catalytic cracking of HDPE which resulted in a wider molecular
mass distribution of products than when zeolite was used [194]. Cracking HDPE
using aluminosilicates results in significantly more cracking compared with pure
thermal pyrolysis, with very high resulting olefin yields, but little increase in aromatic
compounds [190, 195].
2.6.3.3 FCC Catalysts
Fluid catalytic cracking (FCC) catalysts are used commercially by the petrochemical
industry to crack long-chain hydrocarbons into more useful liquid fuel products. These
are generally proprietary mixes of aluminosilicates and zeolites, with a resulting cross-
section of porous structure, and are highly effective at step-wise cracking, first with
heavy molecules in mesopores, then lighter ones in micropores [13]. Given their indus-
trial use, it is not surprising that these catalysts have been shown to be more selective
for liquid products rather than gases when employed for pyrolytically cracking HDPE,
with the products containing a greater proportion of olefins, aromatics/paraffins, and
branched species than thermal cracking alone [189, 195, 196].
2.6.3.4 Activated Carbon
During the course of this study, González et al. [197] conducted an investigation in
which activated carbon, along with zeolite and a silica acid-catalyst were used to crack
polyethylene in an electrically heated fixed bed reactor. The highest yield of products of
49.2% of input mass was obtained with activated carbon at 450°C. At this temperature
the products were comprised of aliphatic species in all cases except with the activated
carbon, where the condensed products also contained around 6% aromatics. This study
used only small amounts of catalyst (a 1:10 ratio of catalyst:plastic), and exhibited
across-the-board low conversion rates from plastic into products.
40 Literature Review
2.6.3.5 Summary of Catalytic Cracking
The catalytic cracking of plastics shares a number of common features; it is able to
occur at lower temperatures than pure thermal cracking, and for a given temperature
proceeds at a more rapid rate, resulting in greater cracking than in the non-catalytic
case. As a result of the greater cracking a more narrow range of lighter products is
produced, with a greater gaseous fraction. The primary pyrolysis products resulting
from the catalytic cracking of HDPE are olefins, with aromatics produced in much
greater quantities than thermal cracking in many cases. As a result of the different
catalytic reaction pathways, isomerisation and branching are more prevalent than
with thermal cracking. Overall, the physical and chemical properties of the catalyst
determine the activity and selectivity with which the reaction proceeds.
2.7 Summary
Activated carbon is a polycrystalline form of carbon with a very high level of porosity
and adsorptive ability. These properties arise from a complex three-dimensional
structure that is formed as a result of defects that alter the carbon lattice away from
the regular graphene formation, with pores existing as the interstices. The activation
process, which involves either gasification at high temperatures, or the addition of
chemicals, creates new porosity, widens existing porosity, opens access to previously
inaccessible pores, and modifies the surface of the carbon, to its final highly porous
configuration.
Activated carbon is currently used in a wide range of applications that revolve
around its highly developed porous structure, such as treatment and purification,
product recovery, and the enhancement of the catalytic ability of other substances by
increasing the interface between catalyst and substrate over a wider physical area.
Activated carbon has also shown promise as a catalyst in its own right in reactions
involving hydrogen, carbon, and combinations thereof (including the decomposition
of short-chain hydrocarbons such as methane) which indicate that it may have the
potential to act as a catalyst in the pyrolysis of plastics.
The pyrolysis of plastics has been carried out by a number of researchers, producing
lighter hydrocarbon products that have potential value as a fuel or chemical feedstock.
The pyrolysis of HDPE produces primarily linear aliphatic species, with aromatics
becoming more prevalent at temperatures above 600°C. The pyrolysis of HDPE using
acid-catalysts allows the reaction to proceed at a lower temperature with greater
2.7 Summary 41
cracking, an increase in aromatic and olefin yields, and species containing a greater
level of branching.
Microwave-assisted pyrolysis has been used to process materials such as hazardous
and problematic waste, as well as various kinds of biomass and hydrocarbon-based
materials. In these applications it has shown several advantages over pure thermal
pyrolysis such as higher yields, faster reaction time, less emission of harmful undesir-
able products, and reduction in overall energy consumption. Whether these results
arise from the characteristics of microwave heating (fast heating rates, volume heating,
high power density, material selectivity, even heating, and good heating control) or
microwave-specific effects over and above these (such as from direct interaction with
functional groups or alignment of molecules) is the subject of some debate.
The microwave-assisted pyrolysis of plastic is currently poorly studied, though its
possibility has been successfully demonstrated with the use of a carbon bed acting
as a microwave-receptive intermediate. Activated carbon has also seen limited study
as a catalyst in its own right, and has not before been been used in combination
with microwave heating. Given the promise that pyrolysis has shown as a recycling
technique for plastic, the advantages of microwave heating, the requirement of using
a microwave-receptive intermediate when targeting microwave-transparent materials
such as plastics, and the catalytic potential that activated carbon has shown in
hydrocarbon reactions, there is a clear opportunity to combine these by investigating
the microwave-assisted pyrolysis of plastic using an activated carbon bed, and it is this
topic that the remainder of this dissertation comprehensively examines.
Given the paucity of literature in this specific area (this being the first study
to combine activated carbon as catalytic bed with the microwave-assisted pyrolysis
of plastic), the initial focus of this research is on demonstrating the feasibility of
the concept, i.e., developing the necessary experimental apparatus to carry out the
process, and employing it for proof of concept studies. This will allow a comparison
to the existing body of microwave-assisted pyrolysis literature, an evaluation of the
catalytic abilities of activated carbon in this context, and an evaluation of the chemical
composition of the produced products. Assuming these initial results are sufficiently
promising the ultimate goal of this project is to further the understanding of the
microwave-assisted pyrolysis of plastic, and the use of activated carbon as a catalyst,
with a view towards developing a sustainable process that could be employed in an
industrial context.

3
Development of Microwave-Assisted
Pyrolysis Equipment
3.1 Introduction
Many of the key pieces of microwave equipment used in this work existed in the lab, and
had been used for exploratory work involving the pyrolysis of liquid hydrocarbons some
years prior to this work. However, considerable time and resources were invested into
developing, refining, and adding to the apparatus, so that it was suitable for processing
solid HDPE, and moreover ensuring that safe, reliable, and optimal operation was
possible. This chapter describes the equipment, as received, after assembling and
recommissioning, and outlines the most significant developments and improvements to
the equipment undertaken throughout the course of this research.
43
44 Development of Microwave-Assisted Pyrolysis Equipment
to extraction
system
1
H2O
H2O
H2O
2
3
4
5
6
7
8
9
1011 12 13
14
15
16
17
Figure 3.1: Microwave-assisted pyrolysis apparatus as received. Numbered components are
described in the text.
3.2 Description of the Apparatus as Received
Figure 3.1 shows the microwave heating equipment and reactor as received at the
commencement of this research. The reactor vessel itself (1) was a custom-built
cylinder constructed of cast iron, with an internal diameter of 250 mm, a height of
250 mm and a wall thickness of 25 mm. The reactor contained a stirrer (2) that
measured 80 mm from the bottom of the reactor, and extended across its entire width.
The stirrer blades slanted at 30° from vertical (to ensure vertical mixing as well as
radial), and it rotated at a variable speed from 0–10 r.p.m. The reactor possessed two
inlets, one 2 mm I.D. inlet (3) suitable for pumping liquids in, and a larger 40 mm I.D.
inlet (4) sealed with a ball valve fitting (5). Four K-type iconel-sheathed thermocouples
(6; RS Components) were distributed throughout the inner surface of the reactor wall
and were connected to digital read outs (7; RS Components) to monitor the temperature
of the reactor contents. Gases generated in the reactor exited to a Liebig condenser
(8); condensate collected in a measuring cylinder below (9), and non-condensable gases
vented to the extraction system.
The microwave heating system consisted of a magnetron head (10; MH3000S-
3.3 Maximising Transmission of Microwave Power to the Reactor 45
210BA, Muegge) that operated at at 2.450 GHz (+/- 10 MHz) and was driven by a
3 kW air-cooled power supply (11; MS3000D-110EF, Muegge) that could be adjusted
by hand to a power level between 10–100%. An isolator (12) prevented any energy that
was reflected back from the reactor from returning to (and damaging) the magnetron;
the reflected energy was safely diverted into a continuously supplied water load, and
the magnitude of the reflection was continuously measured so that the effective power
output (i.e., that reaching the reactor) at any given time could be calculated. The
magnetron and isolator were connected to the reactor via a series of IEC R26 / EIA
WR340 standard waveguides, including a 90° H-bend (13) (bent in the plane of the
magnetic field component of the electromagnetic wave), which was necessary in order
for the magnetron head and isolator to be physically supported (by a lab jack (14)),
while directing the microwaves into the vertical waveguide port in the reactor lid.
The microwave heating system was hermetically separated from the reactor by a
quartz window (15), which prevented pyrolysis products from entering and damaging
the sensitive electromagnetic components, while still allowing microwaves to pass
through to the reactor. The window was cooled by a water jacket (16) below. A gas
inlet (17) allowed nitrogen to enter the reactor to purge any oxygen to maintain an inert
atmosphere throughout operation. Gas-tight seals between components were achieved
with a milled groove in the flange of the waveguide pieces in which a graphite gasket
was located, with each join placed under sealing pressure by 10 M5 bolts.
3.3 Maximising Transmission of Microwave Power to
the Reactor
One of the main concerns identified when using the equipment, as received, to perform
initial feasibility testing of the pyrolysis of HDPE, was the high level of microwave
energy emitted from the magnetron that was not being absorbed by material in the
reactor. This was due to high amounts of energy being reflected back from either
the reactor itself, or components between the magnetron and the reactor; or, by
components between the magnetron and the reactor absorbing energy before it could
reach the reactor. Power that was reflected back to the magnetron was measured by a
microwave detector diode (MM1001B-110AB, Muegge), and then diverted into a water
load by the return isolator, preventing it from reaching and damaging the magnetron.
When first tested, the measured reflected power under normal operation was 50–60%,
46 Development of Microwave-Assisted Pyrolysis Equipment
an inefficient and dangerous condition that required rectifying before pyrolysis work
could commence.
3.3.1 Arc Prevention
When dealing with high powered electromagnetic fields such as the microwaves used
for heating in this work, there exists the risk of arc formation, where high voltage
ionises the normally non-conductive air across a voltage gap, causing a current to
flow, and effectively creating a “short-circuit”. The very high temperature plasma
that results in this arc is highly destructive, capable of melting or vaporising most
materials. Unchecked, an arc will cause pitting damage, and travel up the waveguide
back towards the power source; if it reaches the magnetron it will short and shatter this
sensitive and expensive piece of equipment. Microwave energy not dissipated in the
reactor is spurious energy that might potentially start an arc, providing an additional
motivation to ensure that as much of the emitted energy was transmitted to the reactor
bed as possible.
3.3.2 Minimising Energy Lost in Transmission to Reactor
3.3.2.1 Locating Energy Loss Source
In the first instance, components of the pyrolysis rig (other than the reactor) that
were absorbing microwave energy were located by gradually ramping up the power
output of the magnetron, and feeling by hand areas that increased in temperature.
This soon revealed that the primary location absorbing power outside the reactor was
the 90° waveguide H-bend (see figure 3.1). In spite of the bend being an off-the-
shelf component, the changes in the magnetic field required to rotate the microwave
beam were causing reflections, and resulting in a significant amount of energy being
absorbed by the bend itself. The high levels of energy reflected back from further
downstream in the microwave transmission system may have resulted in constructive
interference occurring, exacerbating the situation further.
3.3.2.2 Simplifying the Power Transmission System
Several possibilities were considered to overcome the reflection and absorption issues
presented by the waveguide H-bend. For reasons of time and economy, it was decided
to focus on the magnetron head support system to see whether an alternative design
could be employed to eliminate the bend entirely. With the goal of simplifying the
3.3 Maximising Transmission of Microwave Power to the Reactor 47
number and type of components in the system, every piece between the magnetron
and the reactor was evaluated as essential to operation or not, and those regarded
as extraneous were set aside. Discarding the spacing and routing waveguide sections
resulted in a reduction of the transmission path length between the magnetron and
reactor from 1200 mm to 400 mm. With only the core components of the transmission
path (magnetron head, isolator, quartz window, water jacket, and gas inlet) remaining,
a new approach to supporting the magnetron head was tried, in which the components
of the transmission path themselves acted as supports in a vertical stack; with a full
complement of ten M5 nuts and bolts at each join, this approach indeed proved to be
feasible.
The elimination of the waveguide bend, and the simplification of the power trans-
mission pathway, served to reduce the reflected microwave power to around 30–40%.
Sustained testing at maximum power in this linear configuration identified no further
components in the transmission pathway absorbing microwave energy, and it was
subsequently assumed that all energy emitted from the magnetron was either absorbed
in the reactor, or reflected back to the isolator where it was measured by the detector
diode.
3.3.3 Window Development
In order to determine the contribution the remaining components in the microwave
transmission path made to the reflected power, each component was removed in turn
and the pyrolysis rig was run without the input of any sample. The only significant drop
in reflected power (of 10–15%) was observed when the quartz window (separating the
dirty/hot section of the pyrolysis rig from the clean electromagnetic components) was
removed from the transmission path. In addition to causing a substantial portion of the
microwave power reflection, the existing window had several other problems, many
of which were a result of the fact that the design consisted of a number of separate
physical and functional components; these are outlined below:
• Every segment of the window required a graphite gasket in order to maintain a
gas-tight seal; this not only made it more difficult to maintain a good seal, but the
distinct materials presented barriers to electrical continuity, acting as potential
sources of arcs and reflections.
• The multiple pieces of the window were difficult to dis- and re-assemble — when
assembled they had to be in exact alignment along the inner surface in order to
48 Development of Microwave-Assisted Pyrolysis Equipment
act as an effective waveguide and prevent arcs and reflections occurring at the
junctions.
• The water-cooling jacket was physically separate from the window-surround,
meaning the window itself was subject to sub-optimal cooling. This could lead
to heat damage and sometimes even shattering of the quartz window, especially
when deposits built up, making it more microwave-absorbent.
• The quartz windows themselves were very expensive to replace, requiring custom
fabrication, with correspondent long manufacturing lead-times.
3.3.3.1 Miniaturisation: Integration of Water Cooling and Gas Inlet
A redesign of the window (figure 3.2) was undertaken to remedy the problems outlined
above. Given the opportunity, it was decided to incorporate as many pieces of the
transmission pathway as practicable into one unified structure. As such, a new piece
was designed and fabricated, incorporating the window support, internal channels for
water cooling, and a four-point inlet structure for nitrogen purge gas to enter the
reactor. This design was considerably more compact, and was made of one unified
milled aluminium piece, ensuring a smooth internal structure throughout, dramatically
reducing the possibility of arcs forming, and ensuring significantly better performance
of the water cooling.
3.3.3.2 Investigating Alternative Window Materials
Testing with new window support in place, but with the actual quartz glass removed,
indicated that it was the quartz itself, and not the support structure that was the major
cause of microwave reflection, implying that it was not as microwave transparent as
originally assumed. Given the expensive and sometimes fragile nature of the quartz,
an extensive search for a suitable replacement window material was undertaken, with
the requirements that it be microwave transparent, robust under high temperatures,
electrically insulating, and preferably inexpensive. After testing a range of materials
including HDPE, PTFE (Teflon), PVC, rubber, and borosilicate glass, it was eventually
determined that silicone polymer exhibited the greatest number of desirable properties:
it is thermally stable up to 300°C, chemically resistant, almost completely microwave
transparent, an insulator, inexpensive, and does not reflect microwave energy in the
way the quartz window did. Furthermore, on a practical level, using silicone as
the window material eliminated the need to have sealing O-rings on either side of
the window: the compression provided by making the window slightly thicker than
3.3 Maximising Transmission of Microwave Power to the Reactor 49
Figure 3.2: Final design of the microwave-transparent window, which separated the microwave
heating system from the reactor and its associated heat and harmful pyrolysis products. The
mica and silicone sandwich window (1) slotted into the milled aluminium window support (2),
with the silicone providing a good seal due to compression from the window support cover (3).
Water travelled from external inlet (4) to outlet (5), through internal passages in the window
support walls, providing cooling. Nitrogen purge gas entered through an external swagelok inlet
(6) and was distributed to outlets on each of the four internal walls (7).
the space in the window support structure allowed the window itself to maintain an
excellent seal.
3.3.3.3 Reduced Carbon Deposition from Silicone Material
An additional benefit to using silicone as the window material revealed itself after
prolonged use, when it was noticed that significantly less carbonaceous deposit was
present on the window after a number of pyrolysis tests. These deposits generally
accumulated over a number of runs, and arose from cracked hydrocarbon settling
or condensing on the window, where, directly in the path of the microwave beam, it
would be carbonised owing to the high energy flux present. This presented a problem,
as the carbonaceous material itself was highly microwave absorbent and, without
regular cleaning, the energy absorbed by the build-up would melt or crack the window
itself. The silicone window, however, if not actively repulsive to hydrocarbon, was not
conducive to hydrocarbon settling, and use of this material reduced the carbonaceous
build-up problem significantly.
50 Development of Microwave-Assisted Pyrolysis Equipment
3.3.3.4 Final Window Design Refinements: Added Strength from Central
Mica Layer
A final enhancement to the silicone design was made some time later, when mica
was trialled as an alternative window material. This pseudo-hexagonal crystalline
sheet silicate mineral exhibits excellent electrical insulation properties and microwave
transparency, and remains structurally rigid and thermally stable up to 900°C. The
added strength and hardness of the mica provided an additional backup in the event
that arcing or high temperatures did occur: silicone by itself would decompose
relatively rapidly, allowing damaging pyrolysis gases into the sensitive electromagnetic
components; mica gave an additional safety factor and was much more resistant to
decomposition. However, mica did not exhibit the same resistance to carbonaceous
deposits as silicone, and required an extra sealing system as it could not make gas-
tight joins with the support structure in the way that silicone could. Combining the
best of both materials was possible by using sheets of silicone on the top and bottom
of a central mica layer. This final “mica-sandwich” design (see figure 3.2 on the
preceding page) was strong, gas-tight, microwave transparent, all-but-impenetrable
even to sustained electrical arcs, resistant to carbon deposition, economical, and easy
to maintain owing to the cheap and disposable nature of the silicone layers.
3.3.4 Maximising Power Transfer through Impedance Matching
Impedance is the property of a component to oppose to the passage of current when
an alternating voltage (such as a high frequency microwave source) is applied. In
microwave heating, the impedance of the source and the load (i.e., the carbon bed
in the reactor) are particularly important — the better the impedance of the load is
matched to that of the source, the better the power transfer from source to load. A short
illustrative background on why this is the case is given in the following section, followed
by a description of the steps taken to optimise the power transfer of the microwave
pyrolysis rig.
3.3.4.1 Background: Impedance Matching Theory
Consider a simplified circuit (figure 3.3) in which the microwave source, of voltage
VS, feeds a load of impedance ZL, and the combined impedance of the source and
transmission pathway is ZS. ZS and ZL effectively form a voltage divider circuit, which
gives the voltage across the load, VL (equation 3.1).
3.3 Maximising Transmission of Microwave Power to the Reactor 51
ZS
ZLVS
Figure 3.3: Simplified circuit used to illustrate the impedance matching principle. VS is the
source voltage, ZS is the combined impedance of the source and transmission pathway, and ZL is
the load impedance.
VL = VS .
ZL
ZL + ZS
(3.1)
Substituting into the formula for power P in terms of voltage and impedance
(equation 3.2) gives the power dissipated in the load, PL (equation 3.3).
P =
V 2
Z
(3.2)
PL =
VS
2.ZL
(ZL + ZS)2
(3.3)
Considering the power transferred to the load as a function of the load impedance
(figure 3.4 on the next page), it is clear that the maximum power transfer load occurs
when the load impedance matches the source impedance.
This can also be shown analytically — differentiating equation 3.3 and setting this
to zero gives equation 3.4:
dPL
dZL
=
VS
2.ZS
(ZL + ZS)3
− VS
2.ZL
(ZL + ZS)3
= 0 (3.4)
i.e. ZL = ZS (3.5)
In reality a more complex wave-based analysis would be required for a fully
descriptive model of the microwave pyrolysis setup, where waves have the potential
to be reflected or transmitted at the junction of each component (or indeed at any
given material change within a component), and the standing waves that result would
produce a variable voltage along the transmission line; however, such an analysis is
beyond the scope of this document and, in any case, the principle illustrated here of
greatest power transfer occurring with matched impedances nevertheless holds true.
52 Development of Microwave-Assisted Pyrolysis Equipment
0
0.5
1
1.5
2
0 1 2 3 4 5 6 7 8 9 10
N
o
rm
a
li
s
e
d
 P
o
w
e
r,
 V
o
lt
a
g
e
, 
C
u
rr
e
n
t 
ZL / ZS 
Power
Voltage
Current
Figure 3.4: Power transferred to the load as a function of the load impedance. Load impedance
is normalised to the source/transmission impedance. Load voltage and current are also shown
for illustrative purposes. Maximum power transfer occurs at ZL
ZS
= 1, i.e., ZL = ZS where the load
impedance is matched with that of the source and transmission path.
3.3.4.2 Tuning the Microwave Circuit
It is possible to alter the impedance of the load through the use of a stub tuner. This
device inserts stubs of capacitive material into the waveguide, at various positions
along a wavelength, close to the load. These stubs act as electrical shunts, and alter
the combined impedance of the tuner and load. Power not dissipated in the load is
reflected back to the source where it is diverted by the isolator. By measuring the
reflected energy and altering the impedance of the load it is possible to “tune” the
circuit so as to minimise reflected energy and maximise power transfer to the load. To
this end, a three-stub tuner (MW2009A-260ED, Muegge) was acquired and added to
the microwave transmission pathway immediately before the window. Appropriately
set through manual experimentation, the tuner was able to reduce the overall reflected
energy; however, the reflected energy was not static over time, making it difficult to
optimise.
3.3.4.3 Time-Domain Impedance Optimisation
In order to gain a greater understanding of the variation in reflected power, software
was written (see §3.6) to sample the reflected energy from the detector diode and track
3.3 Maximising Transmission of Microwave Power to the Reactor 53
Figure 3.5: Microwave energy reflected back from the reactor, exhibiting exactly twice the
frequency as that of the carbon bed stirrer.
it over time. This revealed that the variation in reflected power was periodic, and
semi-sinusoidal (see fig 3.5).
It was soon noticed that the period of the reflected power function was exactly
half that of the period of the carbon bed stirrer. The connection between the two
was determined when the reactor bed was examined with the stirrer running: the
motion of the stirrer created a crest in the carbon that rotated along with the stirrer
(see figure 3.6 on the following page). The microwaves incident to the reactor were
polarised along a directionally constant axis, but the amount of carbon present along
this axis (and thus the impedance of the load) varied as the crest of the carbon induced
by the stirrer rotated.
While it was possible to adjust the tuner to minimise reflected power for any specific
static configuration when the stirrer was stopped, this resulted in an overall much
higher reflectance when the stirrer was running. Instead, by charting the reflected
power over time it was possible to tune the rig such that the area under reflected
power curve was minimised, optimising for the average state of the load impedance.
Using this technique, and the improvements described in the previous segments of
this section, it was possible to reduce the reflected power to an average of less than
10%. Fully optimised, even the peak of the reflected power curve remained well within
acceptable and safe operating limits, ensuring that by far the majority of the microwave
power was successfully transferred to the carbon bed.
54 Development of Microwave-Assisted Pyrolysis Equipment
Figure 3.6: Crests in the carbon bed were induced by the stirrer as it rotated, causing periodic
variation in load impedance and microwave power reflectance (left); in this image the volume of
carbon bed has been reduced until the top of the stirrer is just visible. For clarity, an abstract
approximation of the form of the top surface of the carbon bed induced by the stirrer is given at
right, with the average bed height denoted by the checked grey plane.
3.4 Sample Injection System
Prior to this work, the microwave pyrolysis rig had been used only for processing
liquids, which were injected directly into the carbon bed through a small side inlet.
In order for HDPE to be processed, a system had to be developed that could inject solid
pellets into the reactor under operating conditions. Requirements for the injection
system included that it be gas tight, microwave impenetrable, and temperature
resistant.
3.4.1 Plunger Injection
A simple mechanical plunger apparatus was constructed (figure 3.7), with a stainless
steel tubular barrel of the same internal diameter (40 mm) as the top inlet of the reactor
lid. The plunger fitted precisely inside the inner diameter of the barrel, and had a hand-
operated shaft that was sufficiently long to allow the plunger head to extend all the way
into the reactor itself. This ensured that all of the sample mass entered the reactor, that
no blockages occurred, and that sample pellets did not stick to the inlet tube walls. In
order to operate, the ball valve was closed, and a measured mass of HDPE pellets was
added to the barrel. The plunger was then fitted and lowered to the level of the HDPE
pellets so that minimal air was injected at the same time as the sample. A gas-tight
3.4 Sample Injection System 55
Figure 3.7: Batch addition plunger injection apparatus ready for injection. The PTFE cap at the
top end is screwed down creating a gas-tight seal once the sample has been added; an O-ring
and ball valve seal the bottom of the system until the sample is injected into the reactor.
seal was ensured with a silicone O-ring at the ball valve end of the plunger casing, and
a PTFE cap that screwed on at the open end of the barrel after the sample and plunger
were added. When the reactor reached the desired temperature the ball valve was
opened, the plunger was quickly extended into the reactor and withdrawn, and the ball
valve closed behind it.
3.4.1.1 Problems with Plunger System
The plunger injection system proved invaluable in providing initial exploratory results
for the microwave-assisted pyrolysis of HDPE using coke and activated carbon reactor
beds (see chapter 4). However, a number of problems soon became apparent. The
plunger system was only able to add a maximum of around 100 g of HDPE to the reactor
for any given experiment. This mass was limited by the length of the plunger barrel
— given the small diameter of the injection port, the length of the barrel and plunger
handle would quickly become unwieldy were the mass of the injected sample increased.
The plunger system was only able to inject sample instantaneously, en masse. Not only
56 Development of Microwave-Assisted Pyrolysis Equipment
580
585
590
595
600
605
00:00 05:00 10:00 15:00 20:00
T
e
m
p
e
ra
tu
re
 (
°C
) 
Time from Sample Injection (mins) 
Figure 3.8: With a target reactor temperature of 600°C, the microwave heating system was
unable to maintain operating temperature when 100 g of HDPE pellets were injected, leading to
an 18°C drop in temperature lasting for several minutes.
was this batch addition not a good approximation to any inevitably continuous process
that might be developed were the system to be scaled up, but the quantity of mass
entering the reactor over such a short period of time could overwhelm the ability of the
microwave heating system to maintain the reactor bed at the desired temperature (see
figure 3.8), making the characterisation of the products at a given process temperature
less reliable than desired.
3.4.2 Continuous Rotary Feeder
A continuous feeder was designed, based around a rotary airlock; HDPE pellets would
gravity feed from a 12 L inverted pyramidal hopper into a constantly rotating eight-
spoke feed wheel. The rotational speed of the feed wheel determined how quickly
pellets were transferred from the hopper into the reactor’s top inlet, where they would
fall under gravity into the reactor itself. A Parvalux SD1BSS DC shunt-wound motor
was combined with a worm gearbox in order to provide a sufficiently low revolutionary
speed for the feed wheel. This reduction also provided sufficient torque to shear HDPE
pellets in the event that they became stuck between the feeder spokes and the casing,
ensuring the feeder did not become immobilised during operation. The quantity of
pellets remaining in the hopper could be monitored through an air-tight polycarbonate
3.4 Sample Injection System 57
0
200
400
600
800
1000
1200
1400
1600
00:00 05:00 10:00 15:00 20:00 25:00 30:00
M
a
s
s
 (
g
) 
Time (mins) 
Speed 10
Speed 8
Speed 6
Speed 4
Speed 2
0
10
20
30
40
50
60
0 2 4 6 8 10
G
ra
m
s
 p
e
r 
m
in
u
te
 
Feeder Speed 
Figure 3.9: Calibration of the continuous HDPE pellet rotary feeder. Feeder throughput was
constant and accurate over time (top); furthermore, feeder throughput scaled linearly and
accurately with the feeder motor controller settings (bottom). Coefficient of determination, R2,
was greater than 0.999 in all cases.
view port on the top face of the hopper. The diameter of the feed tube was much less
than the wavelength of the microwaves, effectively isolating the feeder mechanism and
preventing any microwave radiation from leaving the reactor.
3.4.2.1 Calibration of Rotary Feeder
The feeder motor controller operated only in discrete states, with speeds 0–10
available. With no existing information on the output rotational speed at the various
motor settings, a comprehensive calibration of the feeder was required over the full
range of feeder speeds (see figure 3.9). With this calibration it was possible to calculate
an appropriate running time for any desired mass of sample to be introduced to the
reactor, at any feeder speed.
58 Development of Microwave-Assisted Pyrolysis Equipment
3.4.2.2 Co-addition of Air with HDPE Pellets
With the necessity to maintain an oxygen-free atmosphere within the reactor for
pyrolysis to occur, there was some concern that air from the hopper would travel
into the reactor along with HDPE. As such, the hopper was designed and fitted with
vacuum and purge gas feeds, so that it could be purged with a vacuum pump and back-
filled with nitrogen gas before the feeding process began. In practice, experimentation
showed that these steps were unnecessary, and no significant quantity of air entered the
reactor with the HDPE pellets — no oxygenated species were observed in subsequent
GC/MS analysis of products created under these conditions. It is thought that the
volume between HDPE pellets was so minimal as to be insignificant, and what little air
was present may have been prevented from entering the reactor owing to the positive
pressure from the nitrogen purge gas feed, and gas production reactions, within the
reactor itself. Correspondingly, the air-lock nature of the feeder minimsed the return of
pyrolysis gases to the hopper, which was in any case gas tight itself, and thus prevented
any pyrolysis gas escaping to the external environment.
3.4.2.3 Pellet Adherence to Reactor Inlet
One problem occasionally encountered when using the continuous feeder was the
adhering of pellets to the inlet tube downstream of the feeder, where sticky pyrolysis
products would sometimes condense. In the event that this did occur, it usually involved
a minimal amount of mass relative to the total added. Pellets that did not enter the
reactor were recovered at the end of a run and subtracted from those calculated to
have been added. Regular cleaning of the inlet minimised this occurrence.
3.5 Pyrolysis Product Condensing Apparatus
The performance of the condensing apparatus acted as the primary limiting factor
in collecting non-gaseous pyrolysis products, and a number of design iterations were
undertaken in an attempt to maximise the condensed pyrolysis yield.
3.5.1 Liebig Condensers
Initial testing was carried out with a single Liebig condenser, which was present in the
rig configuration as received (figure 3.10a). It quickly became clear that the rapid
production of pyrolysis products when HDPE was added to the reactor completely
3.5 Pyrolysis Product Condensing Apparatus 59
a) b) c) d)
Figure 3.10: Configurations of several tested designs used to condense pyrolysis products as
they exited the reactor. a) Single Liebig condenser. b) Four Liebig condensers in series. c) Three
Liebig condensers with a mesh particle trap. d) Three Liebig condensers with a dry ice / acetone
cold trap. Arrows mark the entry of pyrolysis gases, and the exit of non-condensable gases to the
extraction system.
overwhelmed this single condenser, resulting in a low yield of condensed products,
and necessitated finding a more effective condensing solution.
3.5.1.1 Multiple Condensers
Additional Liebig condensers were added in series, up to a maximum of four (figure
3.10b), with the condensing system tested after each addition. No extra condensed
pyrolysis product was collected when a fourth Liebig condenser was added, so it was
removed in subsequent design iterations. With three condensers in place the system
showed a substantial improvement in performance; however, a significant portion of the
pyrolysis products that could be condensed were still passing through the system. This
was clear both from the yields, which at 35% were still significantly below that observed
in literature [29], and from the presence of a visible white aerosol mist, comprising fine
particles of uncollected pyrolysis product, that exited the condensing system.
3.5.1.2 Temperature Variation
Using recirculating water heater/chiller units, a number of different operating tem-
peratures for the condensers were investigated to increase the yield of condensed
product. In accordance with the findings of Ludlow-Palafox [46], the temperature of
the condensers was raised to 60°C to aid collection. With further experimentation
additional improvements were made; the optimal configuration was found to be 60°C
60 Development of Microwave-Assisted Pyrolysis Equipment
for the first condenser, 25°C for the second, and 0°C for the third. Overall, however,
the condenser temperature had a relatively small effect on yield, making a difference
of only a few percent at most.
3.5.1.3 Attempts to Collect Aerosol Mist Using a Particle Trap
A significant portion of the pyrolysis products exiting the reactor condensed into fine
aerosol droplets, which were light enough that they were swept along by the continuous
flow of gas exiting the reactor. This meant that they travelled down the centre of the
condenser tube and did not make contact with the cold condensing walls. In order
to capture these particles and prevent them from leaving the condensing system, a
mesh of steel wool was introduced at the exit of the Liebig condensers (figure 3.10c),
the idea being to disrupt the laminar flow of the pyrolysis gas and present a greater
cooling surface area onto which any particles could settle. Unfortunately, despite
experimenting with several different materials of varying density (non-absorbent fibre
wool, tissue, wire mesh, etc.), the filter material soon became overwhelmed and
clogged with the throughput of the pyrolysis gas; the pressure build-up would then push
the filter out of its holder and the system would become ineffective. While unsuccessful
as an attempt to collect the mist, this apparatus did at least validate the idea that there
was still a significant quantity of product leaving the system in aerosol form.
3.5.1.4 Difficulties with Sample Extraction
When heavier pyrolysis products were produced they would accumulate on the inner
surfaces of the condenser system. Extracting them required gently warming them
with a heat gun until they reached their melting point and ran into the collecting
measuring cylinders below. Using a complex condensing system comprising many
different individual components required joining each of these components together.
The off-the-shelf Liebig condensers and U-bends used in this work employed standard
QVF flange seals between each component. Unfortunately, the QVF supports around
each join could be damaged by the heat from the heat gun required to melt and collect
the pyrolysis waxes; the use of these types of joints prevented access to a portion of
the condensing system until the extremely time-consuming process of disassembly and
reassembly could be undertaken.
3.5 Pyrolysis Product Condensing Apparatus 61
3.5.2 Alternative Condenser Designs
A number of alternative condenser designs were considered in order to increase surface
contact area between the pyrolysis gas and the condenser system; unfortunately, most
of these were deemed impractical for one reason or another. Condensers such as coiled
condensers, which packed a greater contact surface area of condenser into a smaller
volume, presented difficulties with internal access for cleaning and sample extraction.
A cyclone vortex was considered as a possibility to settle hydrocarbon particles caught
up in the gas stream; however, the gas velocity was insufficient to exert adequate
centripetal force to be effective. Introducing a forcing system to increase the velocity
posed issues of altering reaction pressure, to say nothing of procuring a pump with the
ability to process, and not be clogged by, gas containing hydrocarbons.
3.5.3 Dry Ice Cold Trap
A cold trap chilled by a dry ice and acetone bath to -78°C was introduced in an attempt
to enhance the effectiveness of the condensing system (figure 3.10d); the cold trap
incorporated a small spherical vessel that condensed product could collect in, so as
not to block transmission of gases through the wider system. While the cold trap did
effectively increase the condensed yield by several percent, it still provided a smooth
bore through which the pyrolysis gases could travel, and thus did not address the issue
of hydrocarbon particles travelling in the gas stream, preventing them from settling.
3.5.4 Dry Ice Percolating Condenser
A final, successful, design that incorporated low temperatures, high surface area, and
non-laminar pyrolysis gas flow, was created with a condenser that employed dry ice
as a cooling agent, but on the inside surface of the condenser. In this way, pyrolysis
gas entering the reactor would have to pass through the spaces between the dry ice
pellets — an innately tortuous path that disrupted the laminar flow of the pyrolysis gas.
The pellets themselves, approximately cylindrical in shape, with a diameter of 8–9 mm
and a length of 5–20 mm, also provided a very large surface area, maximising heat
transfer from the gas to the condensing system. The condenser itself (figure 3.11 on the
following page) was fabricated out of thermal-shock-resistant borosilicate glass, and
was funnel-like in shape — the initial wide 150 mm diameter entrance provided a large
area for the incident pyrolysis gas to settle over, preventing blockages from occurring
as the condensed hydrocarbon accumulated on the surface of the dry ice pellets (though
62 Development of Microwave-Assisted Pyrolysis Equipment
1
2
3
4
5
6
Figure 3.11: Final design of dry ice percolating condenser. Pyrolysis gases entered from the
reactor through a port on the lid (1), and passed through the dry ice (2), where they condensed,
and were collected in a measuring cylinder below (3). Gas exited through a vertical tail (4), which
ensured maximum heat transfer and condensed yield. The remaining non-condensable gases
were sampled at a monitoring port (5) just prior to entering the extraction system. A pressure
release valve on the lid (6) prevented excess pressure building up in the event of blockages.
Arrows mark pyrolysis gas entry point and exit to extraction system.
it was found that this was somewhat self-regulating as the dry ice sublimated as
the warmer condensate deposited). The condenser then progressively narrowed in
diameter down to 35 mm, at which point a 1 L collecting vessel was attached below
in which liquid pyrolysis product accumulated. From here the condenser extended
upwards with a 500 mm long tail which was also filled with dry ice, and which ensured
maximum contact between the ice and the traversing pyrolysis gas. The lid of the
condenser was sealed by an O-ring and compression seal, and incorporated both the
inlet through which the pyrolysis gas entered, and a pressure release valve, which
would allow any excess gas to vent into the extraction system if the pressure inside the
condenser exceeded one bar.
Having one, large, condenser rather than a system comprised of several separate
units provided several advantages. The size of the condenser, with its entire top
a removable lid, afforded easy access for cleaning, while the single unit provided a
simplified system, with fewer potential points of failure from seals and joins, and made
it ideal for the recovery of longer chain waxes using a heat gun.
3.5 Pyrolysis Product Condensing Apparatus 63
3.5.4.1 Operation
The condenser was filled with dry ice and connected to the outlet of the reactor
immediately prior to HDPE sample being added to the reactor to avoid the dry ice
sublimating while the condenser was not in use. When pyrolysis was under way, gas
passed through and condensed on the surface of the dry ice pellets. Depending on the
melting point of the condensate, small molecule liquids would drip down into a 1 L
measuring cylinder collection vessel, and larger molecules would freeze to the pellets
and gradually make their way down as the dry ice beneath sublimated. No aerosol
particles were observed leaving the condenser (though sublimated CO2 gas was visible)
— the tortuous convection path and high surface area ensured all particles settled in
the condenser. At the completion of the experiment, the dry ice in the condenser was
allowed to sublimate away, leaving any previously frozen pyrolysis products to melt
into the collecting vessel (with the assistance of a heat gun in the case of sufficiently
long-chain waxes).
It is recognised that a dry ice condenser as described would not be appropriate for
use were the process to be scaled up significantly in size and operational throughput.
It does, however, maximise collection in the laboratory context, and provide useful
information on maximum yield expectations for the process applicable on any scale. An
alternative condensing system design that might be more suitable on a larger scale is
outlined in §7.2.6.
3.5.4.2 CO2 Interaction with Condensate
Concerns that the hot pyrolysis products might react with the CO2 in the condenser
to form new species were allayed through a comparison of the products collected
with this condenser and previous condensing systems. While some differences in the
relative proportions of species in the products were noted (e.g., a better retention of
smaller molecules), this was to be expected with the more efficient condensing system.
Significantly, no oxygenated species were present when using the dry ice condenser
other than CO2 in the non-condensable gases, implying that no chemical interaction
occurred. The rapid cooling of the pyrolysis products upon leaving the hot zone of
the carbon bed in the reactor, and the very low temperature of the dry ice, no doubt
ensured this.
64 Development of Microwave-Assisted Pyrolysis Equipment
3.5.4.3 CO2 Contamination in Collected Gas
While the evaporation of the dry ice from the condenser did add a significant quantity
of CO2 to the gas samples collected for later analysis, testing with previous condensers
had convincingly demonstrated that no CO2 was actually generated from the pyrolysis
of HDPE under normal operation. As such, the CO2 peak in GC/MS analysis of samples
collected using the internal dry ice condenser was discarded (with no loss in resolution
for the analysis of the remaining gases), and no CO2 component was included in the
gaseous product composition analyses.
3.6 Control Software Development
As received, control of the magnetron power supply was effected via a rotary poten-
tiometer, which was set by hand and allowed the variation of power from 0–100%.
3.6.1 Problems with Existing Control Setup
The existing manual setup suffered from a number of limitations, including:
• requiring significant time and attention, forcing the operator to constantly adjust
the power at the expense of other important tasks
• performing inconsistently between runs
• unpredictability — if the operator was not paying close attention to the power
output a wide variety of temperatures could result
• no mechanism for recording the power output, or the reflected power
• limited diagnostic output and feedback when errors occurred
• slow human response in the event of a problem
Given the large number of experimental runs required for the work in this project,
a more robust system remedying these flaws was required; as such, a computerised
automatic controller was developed as described in the following section.
3.6.2 Hardware Interface
The magnetron power supply included the facility for control by an external pro-
grammable logic controller. An in-depth analysis of the control mechanism and
thermocouples resulted in the requirements for a computer-based control system, set
forth in table 3.1. Digital input and output was provided by an Audon PCX-4264
PCI card with 32 channels; analog input and output was provided by a Measurement
3.6 Control Software Development 65
Table 3.1: Input and output requirements for computer control of the microwave power supply.
Input and output is defined relative to the computer controller, i.e., input signals are readings
from the power supply or thermocouples to the computer, and output signals are controls sent
from the computer to the power supply.
Analog Digital (24 V)
Input Output Input Output
Power supply monitor (0–10 V) Magnetron power (0–10 V) Ready Remote control on
Reflected power (0–10 V) Magnetron status Filament heating on
Thermocouple 1 (0–50 mV) High voltage supply Magnetron on
Thermocouple 2 (0–50 mV) Safety interlock Reset
Thermocouple 3 (0–50 mV) Failure
Thermocouple 4 (0–50 mV) Magnetron cooling failure
Thermocouple 5 (0–50 mV) Power supply cooling failure
Input failure
Overload
Computing PCI-DAS6014 16-bit PCI card with eight input channels and two output
channels. The 16-bit resolution provided by the analog input board was judged to be of
sufficient depth, when operating in the 50 mV range mode (providing a theoretical
0.76 µV resolution), to read the thermocouples directly without amplification. A
custom-built daughter board was also constructed to handle switching from the low
voltage digital I/O to the 24 V system used by the power supply.
3.6.3 Software Controller
National Instruments’ Labview 8.6 was selected as a programming environment in
which to develop control software for the microwave system owing to its strengths
in rapid development, real-time capabilities, and data input/output interfacing. Fig-
ure 3.12 on the next page shows a simplified schematic of the main control program,
the components of which are described below. A screenshot of the final configuration
of the software interface is given in figure 3.13 on page 67.
3.6.3.1 Analog Input
The analog signals (thermocouple voltages, magnetron power, reflected power) were
read in at a sampling rate of 40 kHz. Digital signal processing was applied to the
signals to reduce background noise (e.g., a notch filter was applied to remove 50 Hz
mains interference). The thermocouple voltages were converted into temperatures
using the NIST PB93-190338 standard [198] (a ninth-order polynomial approximation
with ≤0.05°C error over the measuring range).
66 Development of Microwave-Assisted Pyrolysis Equipment
Thermocouple 
voltages, V
Magnetron 
Power
Reflected
Power
Analog Input
40 kHz, 16 bit
DSP: filtering/
noise reduction
Convert V to 
temperature, T
Average values 
over Δt seconds
Arc detection
Calculate Error
E
t  = T - Ttarget
Target 
Temperature Ttarget
Calculate Integral
It  = It-1  + E
t .Δt
Cap Max Integral
0 ≤ It  ≤ 20
Tunable Parameters:
Proportional Constant K
E
Integral Constant K
I
Derivative Constant K
D
Analog Output
 Send Magnetron 
Power to Power 
Supply
If Error Detected:
Power off Magnetron
Calculate Derivative
D
t  = E
t  - E
t-1
      Δt
Calculate New Magnetron Power
P
t  = (K
E  x E
t ) + (K
I  x It ) + (K
D  x D
t )
Limit Magnetron Power Increase
(Pt  - Pt-1 )/Δt ≤ 3%
Record Parameters:
Time
Temperatures
Power
Reflectance
User Comment
High sustained 
reflected power 
monitoring
Digital Input
Water Cooling
Overload
Interlock
High Voltage
Power Failure
F
ig
u
re
3
.1
2
:
S
im
p
lifi
e
d
P
ID
co
n
tro
lle
r
so
ftw
a
re
sch
e
m
a
tic.
C
o
m
p
o
n
e
n
ts
a
re
d
e
scrib
e
d
in
§3
.6
.3
.
3.6 Control Software Development 67
Figure 3.13: Screenshot of the final configuration of the PID controller software interface.
3.6.3.2 PID Controller
A software proportional-integral-derivative (PID) feedback controller [199] was con-
structed according to equation 3.6.
P (t) = KE .E(t) +KI .
ˆ t
0
E(τ).dτ +KD.
d
dt
E(t) (3.6)
In this system, the power output to the magnetron (P (t)) at any time (t) is calculated
as a weighted sum of the current error (E(t)) — the difference between the current
operating temperature and the reactor target temperature; a “backward looking”
integral term (
´ t
0
E(τ).dτ) which corresponds to the accumulation of previous errors;
and a “predictive” derivative term ( ddtE(t)) — the current rate of change of the error.
The respective contribution of the proportional, integral, and derivative terms to the
final power output is controlled using user-tunable constants (KE, KI , KD respectively).
Applying equation 3.6 to a computer system requires converting the continuous
equation into a discrete form in the time domain, applicable to individual data points.
This was done for the integral term as follows:
68 Development of Microwave-Assisted Pyrolysis Equipment
ˆ t
0
E(τ).dτ = It = It−1 + Et.Δt (3.7)
where It is the current accumulated integral error, which comprises the previous
integral error (It−1), and the error (Et) accumulated over the previous time step (Δt).
The term is initialised to 0 at the commencement of operation (It=0 = 0).
The derivative term was approximated using backward finite differences, as the
change from the error at the previous sampling time (Et−1) to the current error (Et)
over the sampling time (Δt).
d
dt
E(t)≈Et − Et−1
Δt
(3.8)
Combining these gives the overall discretised PID equation for microwave output
power (Pt):
Pt = KE .Et +KI .(It−1 + Et.Δt) +KD.(
Et − Et−1
Δt
) (3.9)
3.6.3.3 Practical Implementation Issues
Integral windup occurs when the integral term accumulates to a very large value as the
reactor temperature differs from the set point for a prolonged period of time (e.g., when
the reactor is heating up at the beginning of an experiment). In order to prevent this
occurring, a hard limit was imposed on the maximum value the accumulated integral
error could take. An absolute value of 20 for this limit was found to provide plenty
of head-room for the integral parameter during normal operation (around the target
temperature), while preventing integral windup from deleteriously affecting operation.
The selection of a suitable time period, Δt, over which to sample the temperature
inputs (and thus error), and between which to update the magnetron power, involved a
trade-off between providing a sufficiently rapid response to changing conditions (e.g.,
the addition of HDPE sample requiring greater power input to maintain the bed at
the target temperature), and introducing instability owing to the inherent time delay
between changing the power input and the resultant change in measured reactor bed
temperature. Manual experimentation with this parameter resulted in the selection of
a five-second update period that proved both responsive and stable.
Translating the boundless PID function into a practical power output required
limiting the absolute value of the output 0≤Pt≤10 so that it fell between the 0–10 V
of the power supply input. It was also considered prudent to avoid large increases in
3.6 Control Software Development 69
power output at each update interval to prolong the life of the magnetron filament.
Without limitation, when the PID was first turned on the power would jump from 0% to
100% instantaneously, potentially damaging the magnetron with the sudden increase
in power. As such, an upper limit was imposed in software with the maximum power
increase confined to 3% per second.
Tuning of the PID constants (KE, KI , KD) was achieved using the Ziegler-Nichols
method [200] initially, then manually optimised for stability and sensitivity of the
system. The fully tuned system exhibited excellent performance, and was able to
maintain the average reactor temperature within ±1ºC of the target.
3.6.3.4 Arc Prevention and Safety Systems
As described in §3.3.1, arcing is a destructive threat present when operating high
power electromagnetic sources such as the microwave reactor used in this work. When
arcing does occur, the key to preventing damage is to recognise that it is occurring,
and to cut power to the power source as quickly as possible. It was found that a
reliable indicator that an arc was occurring was a sudden spike in reflected power,
far beyond the normal levels usually observed. Accordingly, a detection subroutine that
monitored the magnitude of the reflected power was incorporated into the programme;
the routine triggered when it breached a safety threshold, and instantly powered down
the magnetron. Under computer automation, with a 40 kHz sampling rate (i.e., a
0.025 ms response time), monitoring was far more effective, and action taken far more
rapidly, than when it was left to a human operator.
In addition to monitoring high transient reflectance, the software also monitored
for a sustained higher-than-normal average reflected power, which could indicate an
arc occurring deeper in the reactor (as opposed to in the waveguide) or some other
malfunction, and would also trigger a rapid safety shutdown so the cause could be
investigated.
The software also had a number of digital inputs from sensors distributed around
the power supply and magnetron head that would indicate if a fault had occurred (e.g.,
overload, low water cooling rate, power failure, etc.) and would shut down the system
to prevent damage in the event that any of these were triggered.
3.6.3.5 Data Logging
Throughout execution, the software continuously logged operating parameters (current
time, thermocouple temperatures, microwave power output, reflected power). The
70 Development of Microwave-Assisted Pyrolysis Equipment
software interface also contained the facility to record comments from the operator,
which were time-stamped, and recorded in the log file along with the other parameters.
In the event of catastrophic failure the software was designed to prevent information
loss as much as possible by writing data to disk as soon as possible. In the event that
such failure did occur, the software was designed for rapid resumption of controller
duties so as to minimise disruption to the experiment, and was able to append to the
log file starting from the last data point written.
3.7 Safety Systems
The combination of microwave radiation, high temperatures, and flammable gases
presented a unique series of safety challenges, as our beleaguered department health
and safety officer would no doubt attest.
3.7.1 Microwave Safety
British Standard 5175 states that the maximum safe level of power density permissible
outside the shielding of a microwave device is 50 W m-2 at a distance of 5 cm.
Conformance with this specification was tested every time after the pyrolysis rig was
assembled, by gradually increasing the power output of the magnetron and sweeping
the entire rig with a Robin TX90 hand-held microwave leakage detector. Use of industry
standard IEC R26 / EIA WR340 waveguides and flanges for all components in the
microwave transmission path meant that leakage was rarely above background and
the safety standard was never breached.
3.7.2 Preventing Equipment Leaks
Given the potential toxicity of the gases produced during pyrolysis it was important
to have completely gas-tight equipment. A common design approach employed
throughout improvement of the pyrolysis rig was to simplify components, and combine
them where possible, to reduce the number of failure points due to joins and seals.
In spite of this effort, the rig was still tested after each modification and assembly by
perfusing with nitrogen gas, and testing at each seal with a hand-held Restek thermal-
conductivity leak detector for the presence of nitrogen gas in concentrations above
background.
3.8 Final Configuration of Microwave-Assisted Pyrolysis Apparatus 71
3.7.3 Extraction System
Any pyrolysis gases that were not able to be condensed were sampled and then
entered the extraction system. The extraction system totally surrounded the exit
of the condenser, but was not directly coupled to it, in order to avoid applying
negative pressure to the reactor and condensation equipment. This also allowed the
exiting gases to be substantially diluted by the concurrently admitted air, reducing
the concentration of any toxic components, and diluting the hydrocarbons below the
minimum concentration required for combustion.
3.7.4 Additional Safety Systems
Additional safety systems installed throughout the course of this work include:
• An emergency stop button installed within arm’s reach of the operator that would
immediately cut power to the magnetron.
• An audible and visual alarm which would alert the operator in the event of
extraction failure.
• Additional CO2 fire extinguishers — given the criticality of timeliness when it
comes to fire control it was decided to install several supplementary extinguishers,
such that at least one would be within arm’s reach of the operator at all times.
3.8 Final Configuration of Microwave-Assisted
Pyrolysis Apparatus
The final configuration of the microwave pyrolysis apparatus, incorporating all of
the improvements described thus far, is detailed overleaf in figures 3.14 and 3.15 in
schematic and photographic form respectively.
3.9 Summary
A number of problems presented in the microwave-assisted pyrolysis equipment
as received, including high levels of reflected power, transmission losses, and the
formation of damaging arcs. These problems were tackled with an overall design
approach involving simplifying and combining components where possible, to reduce
the number of potential failure points. This resulted in a rethink of the transmission
pathway — linearising it and eliminating a substantial number of components, and a
72 Development of Microwave-Assisted Pyrolysis Equipment
N2
to extraction
system
H2O
H2O
1
2
3
4
5
6
8
9
10
11
12
13
14
15
16
17
18
19
7
20
1 Feeder hopper 11 Carbon bed
2 HDPE pellets 12 Stirrer
3 Rotary feeder mechanism 13 Thermocouple
4 Ball valve isolator 14 Computer control system
5 Magnetron 15 Pressure release safety valve
6 Isolator 16 Condenser
7 Three-stub tuner 17 Dry ice
8 Window 18 Collecting vessel
9 Nitrogen inlet 19 Peristaltic sampling pump
10 Reactor 20 Gas collecting bag
Figure 3.14: Schematic of the final configuration of the microwave-assisted pyrolysis apparatus.
3.9 Summary 73
1 Computer control system 13 Bed collection container
2 Feeder motor controller 14 Pressure release valve extraction
3 Magnetron power supply 15 Condenser extraction
4 Emergency stop button 16 Condenser
5 Stirrer motor controller 17 Collecting measuring cylinder
6 Fire extinguisher 18 Peristaltic sampling pump
7 Control system daughter board 19 Gas bag
8 Feeder hopper 20 Magnetron
9 Rotary feeder mechanism 21 Isolator
10 Inlet isolator ball valve 22 Three-stub tuner
11 Reactor 23 Reactor gas outlet pipe
12 Reactor emptying ball valve 24 Window
Figure 3.15: Final configuration of the microwave pyrolysis apparatus in overview (left), and
close-up on the transmission path (right). Numbered components are consistent between both
images, though may not be present in both cases.
74 Development of Microwave-Assisted Pyrolysis Equipment
redesign of the window — integrating it with the nitrogen gas inlet and incorporating
more effective internal cooling. A re-examination of the window material itself
resulted in changing it from expensive, fragile, and microwave-reflecting quartz, to
a mica and silicone composite, which had the advantages of strength, self sealing,
the elimination of microwave reflection, and a reduction in damaging carbonaceous
deposits. These improvements, combined with tuning the microwave system through
impedance matching, reduced reflected power from greater than 50% to below 10%,
and greatly improved system reliability.
An initial plunger-type feeder allowed HDPE pellets to be added to the reactor at
pyrolytic temperatures; however, the discrete addition of sample limited the quantity
of material that could be pyrolysed at once and overwhelmed the microwave heating
system’s ability to maintain a constant reactor temperature. Correspondently a
continuous rotary feeder was designed and built, which solved these issues, and
allowed the pyrolysis rig to operate effectively indefinitely, and process arbitrary
quantities of input material.
Manual control of the magnetron power output required significant operator time
and attention, and proved inconsistent and slow to respond to damaging arcs occurring
in the system. A computer-based PID controller was developed that automated control
of the magnetron, improved reliability, and was able to detect and respond to arcs and
other faults safely and far more rapidly than a human operator.
The condensing apparatus proved a significant limitation on extracting condensible
pyrolysis products from the gaseous stream exiting the reactor. A number of design
iterations based around Liebig condensers failed to prevent a hydrocarbon-containing
aerosol mist from being carried along by the continuous flow of gases through the
system. A final, effective design passed the pyrolysis gas through a condenser
containing dry ice pellets, which disrupted the laminar flow of the gas and provided
a very high surface area to maximise heat transfer.
While presented as a single block of work, the improvements outlined in this chapter
represent the continuous evolution and refinement of the microwave-assisted pyrolysis
rig over a prolonged period of time. Nevertheless, unless the equipment changes had
no impact on results, or it has been otherwise noted, it is the final form of the microwave
pyrolysis equipment that, along with the protocol developed alongside, was used to
generate the results that comprise the remainder of this dissertation.
4
Batch Pyrolysis of HDPE using an
Activated Carbon Reactor Bed
4.1 Introduction
Activated carbon is a relatively unstudied catalyst, completely so in the context of the
microwave-assisted pyrolysis of HDPE. Having developed a reliable set of equipment
and operating procedures to pyrolyse HDPE using microwave heating and a carbon
bed, it was next important to evaluate the utility of activated carbon as a catalyst in this
context, i.e., to establish the feasibility of using activated carbon as a catalyst in this
process, and to determine the characteristics, chemical make-up, and economic value
of the products resulting from its use, so that the overall process could be assessed as
worthwhile or not.
In order to evaluate the performance of activated carbon as a catalytic bed, a
series of exploratory experiments were undertaken using the reactor and equipment
described in the previous chapter, in semi-batch mode (i.e., with the batch input of
75
76 Batch Pyrolysis of HDPE using an Activated Carbon Reactor Bed
HDPE and continuous removal of pyrolysis gases as they were produced), using the
plunger injector (see §3.4.1), hereafter referred to as “batch” mode. The thermal
sensitivity of the process was characterised with a series of experiments conducted
over the widest possible range of operating temperatures: from the lower threshold
of pyrolysis occurring at 400°C, to the maximum temperature sustainable by the
magnetron with the addition of the HDPE at 600°C, with the resulting yields, mass
distribution, and chemical composition of the products examined. Control experiments
were conducted with the same parameters but using a coke bed to provide a baseline
from which to compare the effects of the activated carbon. While the microwave-
assisted pyrolysis of HDPE using a coke bed has been studied previously [29, 46],
specific outcomes are often sensitive to the exact reactor type and configuration.
Accordingly, it was considered prudent to generate control results with the coke bed
for direct comparison, in order to eliminate as many variables as possible.
This chapter describes the methodology, results, and analysis undertaken in order
to establish the utility of activated carbon for microwave-assisted pyrolysis, and places
this in context with the wider body of pyrolysis literature.
4.2 Materials
4.2.1 HDPE
HDPE was chosen as a model material for pyrolysis over the course of this research,
owing to its simple, pure hydrocarbon molecular structure, which avoided adding any
further complication to the already complex pyrolytic breakdown reaction pathway.
HDPE is also one of the relatively few materials on which microwave pyrolysis research
has already been conducted [29, 46], and thus offered an external point of comparison
to this research. Liten ML 71 HDPE was acquired from Unipetrol RPA; this particular
form of HDPE has a wide variety of applications derived from its intended use in
injection moulding manufacturing. It was chosen for use in this research because of
its lack of additives, wide availability, and ease of handling owing to its pelleted form
(cylinders approximately 2 mm in diameter and 3–4 mm long). The material properties
of Liten ML 71 are given in table 4.1.
4.2 Materials 77
Table 4.1: Typical material properties of Liten ML 71 HDPE. Provided by Unipetrol RPA.
Property Value Test Method
Density 963 kg/m3 ISO 1183
Yield Stress 26 MPa ISO 527
Tensile Strain at Yield 8.50% ISO 527
Flexural Modulus 1100 MPa ISO 178
Charpy Notched Impact Strength, 23°C 6.5 kJ/m2 ISO 179
Charpy Notched Impact Strength, -30°C 5.5 kJ/m2 ISO 179
Shore Hardness 58 ISO 868
Vicat Softening Temperature 126 C ISO 306
Melt Flow Rate (190°C / 2.16 kg) 8.5 g / 10 mins ISO 1133
Melt Flow Rate (190°C / 5 kg) 23.0 g / 10 mins ISO 1133
4.2.2 Coke
High purity FC-250 petroleum coke (Timcal Ltd) was obtained for use as the reactor
bed. The high calcination temperature used in production ensured that it contained
no volatile hydrocarbons, and minimal other impurities, which might contaminate the
pyrolysis products. Typical material properties are given in table 4.2. Approximately
2.0–2.5 kg of coke was used for the reactor bed — sufficient to completely cover the
stirrer while in operation. Before the coke was first used for pyrolysis it was heated in
the reactor to 700°C for 30 minutes to ensure any trace compounds were burnt off, and
to remove any moisture from the material before it was first weighed. Heating the coke
to 100°C higher than it would be used at during pyrolysis provided a safety margin that
ensured no contaminants would evolve during the actual pyrolysis experiments.
Table 4.2: Typical material properties of Timrex FC-250 coke (Timcal Ltd), and Aquacarb 207EA
(Chemviron Carbon); data provided by the manufacturers.
Property Coke Activated Carbon
Ash 0.07 wt% <1 wt%
Moisture 0.02 wt% <5 wt%
Sulfur 1.3 wt% -
Bulk density 0.80 g/cm3 0.48-0.52 g/cm3
Bulk volume 125 mL / 100g -
Tapped density 0.91 g/cm3 -
Plastic deformation 0.21% -
Elastic deformation 0.10% -
Surface area - 950–1100 m2/g
Particle Size Distribution 2% >250 µm 0.42–1.68 mm
75% >125 µm
85% >90 µm
78 Batch Pyrolysis of HDPE using an Activated Carbon Reactor Bed
4.2.3 Activated Carbon
Aquacarb 207EA was acquired from Chemviron Ltd. This high-activity carbon originally
derives from bituminous coal, and is steam activated. Typical material properties are
given in table 4.2. Similar to the coke bed, sufficient activated carbon was used as the
reactor bed to cover the stirrer while in operation (2.0–2.5 kg), and an identical initial
heat treatment process was applied to ensure moisture and impurities were removed.
4.3 Microwave Pyrolysis Method
The microwave-assisted pyrolysis apparatus used in this chapter is the final configura-
tion described in §3.8, with the exception that the continuous feeder had not yet been
developed during these experiments. As such, the reactor operated in batch mode, with
the plunger injection system adding HDPE sample in one quantity, and the pyrolysis
products exiting the reactor to the condensing system as they were produced. All
experiments were conducted at atmospheric pressure.
At the commencement of an experimental run, the nitrogen supply was turned on
and allowed to vent through the apparatus at a rate of 0.7 L/min. The stirrer was set
to a speed of 8 revs/min — a speed sufficiently rapid to maintain an even temperature
distribution throughout the carbon bed. The water cooling system was started, and
the computer was set to the target temperature. The magnetron was switched on, and
the reactor bed was heated to the target temperature, a process that took 1–2 hours.
Once the bed had reached the desired temperature it was allowed to equilibrate for
a minimum of 15 minutes to ensure a homogeneous heat distribution. At this point,
the percolating condenser was filled with dry ice and it was connected to the outlet
of the reactor. HDPE pellets were added to the reactor using the plunger injector
system (described in §3.4.1), and the peristaltic pump which sampled the gas exiting
the condenser was switched on.
The pyrolysis of HDPE produced a mixture of gaseous hydrocarbons, some of
which, once they left the hot zone of the reactor, condensed into a suspension of
liquid droplets. This aerosol was visible entering the condenser, and provided a
visual cue as to the progress of the pyrolysis reaction: the reactor was maintained
at the target temperature until more of the pyrolysis aerosol was generated. At this
point the microwave heating was switched off, and the reactor was allowed to cool.
During this time the dry ice in the condenser sublimed, and any frozen condensed
4.4 GC/MS Analysis 79
sample melted into the collecting cylinder. Non-condensable gases continued to be
collected throughout this time (with the CO2 peak arising from the dry ice condenser
subsequently discarded in later analysis). In the event that any solid pyrolysis products
remained in the condenser after all the dry ice had sublimed it was gently warmed with
a heat gun until it melted into the collecting vessel.
Once the reactor bed reached approximately 150°C (a process usually taking 2–3
hours) the carbon was removed from the reactor and weighed; the reactor bed was
always weighed at >100°C so that moisture absorbed from the air was not included in
the mass.
4.3.1 Experimental Conditions
Pyrolysis experiments were conducted with the activated carbon bed, and a control bed
of coke, at 400°C, 450°C, 500°C, 550°C, and 600°C. These temperatures represented
the effective limits of the pyrolysis equipment used: below 400°C pyrolysis either did
not occur or occurred so slowly as to make experiments impractically long; 600°C
was the maximum reactor temperature that the 3kW microwave heating system could
sustain without an unacceptable drop of temperature occurring when the HDPE was
injected. In all experiments 100 g of HDPE pellets was added to the reactor in each
sample injection.
4.3.2 Pyrolysis Product Mass Measurement
The mass of the condensed pyrolysis products was assessed by weighing the full
collecting cylinder at the conclusion of an experiment, then subtracting the mass of the
empty, clean, and dry cylinder. The mass of the input sample retained in the reactor
was measured by the increase in mass of the reactor bed between runs. The amount of
non-condensable gas generated in the pyrolysis was determined through the principle
of conservation of mass, and was calculated by the mass of sample added, less the mass
of the condensed products, less the increase in mass in the carbon bed.
4.4 GC/MS Analysis
Samples were analysed by gas chromatography coupled with mass spectrometry
(GC/MS) using an HP 6890 gas chromatograph and an HP 5973 mass selective system
80 Batch Pyrolysis of HDPE using an Activated Carbon Reactor Bed
(Agilent Technologies). Compounds were identified using Agilent Chemstation version
G1701DA and the NIST 2005 mass spectral library.
4.4.1 Condensed Product Analysis
Condensed pyrolysis products were analysed with an SGE HT5 30 m x 0.25 mm x 0.1 μm
capillary column (5% Phenyl equiv. polycarborane-siloxane). Any samples containing
solid waxes were heated before being injected to ensure they were entirely liquid and
homogeneous. 1 µL of sample was injected into a split inlet operating at 250°C at a
ratio of 100:1, a pressure of 5 psi, and a gas flow of 1.2 mL/min. The oven was initially
set at 30°C, at which temperature it was held for 3 minutes after injection, and it was
then ramped at 8°C/min to 350°C, where it was held for 30 minutes. 6.0-grade helium
(99.9999% pure) was used as the carrier gas.
4.4.2 Gas Analysis
Gas samples were analysed with an Agilent GS-Gaspro 60 m x 0.32 mm porous layer
open tubular capillary column. 100 µL of gas was injected into a split inlet operating at
250°C at a ratio of 20:1, a pressure of 5 psi, and a gas flow of 1.1 mL/min. The oven was
initially set at 30°C, a temperature at which it was held for 10 minutes after injection,
and it was then ramped at 5°C/min to 150°C. 6.0-grade helium (99.9999% pure) was
used as the carrier gas.
4.4.3 Quantification of Compounds by Peak Area
The large number of individual compounds present in the pyrolysis products (up to
several hundred) meant that quantitative calibration of each of these in the GC/MS
analysis was impractical; instead, a measure of proportional peak area is used
throughout this work. With this metric, each compound is quantified by the integrated
area of the corresponding peak on the mass spectrometer chromatogram. This is then
given as a percentage of the total area of all peaks in the analysis. This method allows
a descriptive and quantitative analysis of products that gives a good approximation to
mass percentage.
4.5 Analysis of Condensed Pyrolysis Oils and Waxes 81
Figure 4.1: Condensed pyrolysis products generated using activated carbon (left) and coke
(right) reactor beds. The activated carbon derived condensate is entirely liquid at room
temperature and clear — notice that it is sufficiently transparent that the sample label is clearly
visible at the rear of the glass sample bottle. The condensate produced using a coke bed, by
contrast, is mostly solid at room temperature and entirely opaque.
4.5 Analysis of Condensed Pyrolysis Oils and Waxes
4.5.1 Visual Assessment
Figure 4.1 shows samples of the condensed pyrolysis products produced with activated
carbon and coke beds. A dramatic visual difference is obvious: the condensate
produced using a coke bed is viscous, waxy, and of mixed consistency; when using
an activated carbon bed a clear, homogeneous product resulted that was entirely liquid
at room temperature.
4.5.2 GC/MS Analysis and Peak Identification
Figure 4.2 on the next page shows typical examples of GC/MS chromatograms of the
condensed pyrolysis products produced using activated carbon and coke reactor beds.
The coke derived chromatogram is similar to those produced by other researchers
investigating the pyrolysis of polyethylene [29, 173, 201]. Of particular note is the
regular series of peaks, with each peak grouping further along the chromatogram
corresponding to the group of hydrocarbons with one additional carbon atom in their
molecular formula. The largest and most clearly identifiable of these peaks is composed
of, for each carbon number, a terminal alkadiene (with the double bonds located at
82 Batch Pyrolysis of HDPE using an Activated Carbon Reactor Bed
2 4 6 8 10 12 14 16 18 20 22 24
0.5
   1.0
 1.5
   2.0
 2.5
   3.0
 3.5
   4.0
 4.5
   5.0
 5.5
   6.0
Retention Time (mins)
Ab
un
da
nc
e (
x1
07 )
5 10 15 20 25 30 35 40 45
0.5
   1.0
 1.5
   2.0
 2.5
Retention Time (mins)
Ab
un
da
nc
e (
x1
07 )
4 5 6 7 8 9 10 11
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Retention Time (mins)
Ab
un
da
nc
e (
x1
07 ) n-alkane
1-alkene
alkadiene
Figure 4.2: Typical GC/MS chromatograms generated from the analysis of condensate from
activated carbon (top) and coke (middle) reactor beds. A magnified view of the characteristic
alkadiene, 1-alkene, and n-alkane triple peaks is shown in the bottom sub-figure; the regularity
of these peaks makes them useful as markers for compound identification.
4.5 Analysis of Condensed Pyrolysis Oils and Waxes 83
the peripheries of the molecule), followed by a 1-alkene, followed by an n-alkane (in
order of increasing retention time); this “triple peak” grouping has been identified
by researchers conducting similar analyses of the condensed pyrolytic products of
polyethylene [46, 173, 191, 201]. These regular triple peaks provided valuable
information that could be used to identify compounds in the event that they were either
not contained in the spectral database, or if the match to the database for a peak was
unclear (generally molecules larger than C25 were not present in the NIST database).
Fortunately, the regularity with which the triple peaks occur (a grouping is present
for every carbon number in the sample) provided a consistent goalpost with which to
measure the carbon numbers of unidentified peaks; furthermore, the order with which
different molecules elute remains consistent between (and within) each triple peak
marker, e.g., the alkane form always immediately follows the 1-alkene, allowing the
identification of the specific hydrocarbon from the relative retention time.
In contrast to the coke bed, the activated carbon derived chromatograms show a
much shorter total elution time, with correspondingly smaller molecules on average.
The regular triple peaks are still present, though they are interspersed with a much
larger number of other (mostly aromatic) compounds. The smaller maximum molecular
size of the condensate produced with an activated carbon bed means that the peaks
could be identified using the NIST database alone. Histograms and statistics derived
from the GC/MS analyses are presented in §4.5.3.
4.5.3 Mass Distribution of Condensed Pyrolysis Products
Figure 4.3 on the following page shows the size of the compounds in the condensed
pyrolysis products as measured by carbon number, and determined by GC/MS analysis.
The dramatic difference in the composition of the products is clearly visible, with the
oil produced using the activated carbon bed consisting of much smaller molecules,
distributed over a smaller range of molecular masses. The distribution of the condensed
products produced using an activated carbon bed is uni-modal, centred around C8. By
contrast the distribution of products produced with a coke bed is less regular, with a far
greater spread of molecular masses that encompasses a second grouping of products
in the C40–C50 range (discussed further in §4.5.3.4). Summary statistics quantifying
average molecular mass and spread across all reaction temperatures are given in
table 4.3 on page 85. The pyrolysis oil produced using an activated carbon bed is far
lighter, with an average carbon number of 10.0 to 11.5 compared with that produced
with a coke bed of 25.0 to 33.0 — approximately three times larger. Furthermore,
84 Batch Pyrolysis of HDPE using an Activated Carbon Reactor Bed
0
2
4
6
8
10
12
14
16
18
5 10 15 20 25 30 35 40 45 50 55
%
 T
o
ta
l 
P
ea
k
 A
re
a
 
Carbon Number 
400°C 
0
2
4
6
8
10
12
14
16
18
20
5 10 15 20 25 30 35 40 45 50 55
%
 T
o
ta
l 
P
ea
k
 A
re
a
 
Carbon Number 
500°C 
Activated Carbon
Coke
0
2
4
6
8
10
12
14
16
18
5 10 15 20 25 30 35 40 45 50 55
%
 T
o
ta
l 
P
ea
k
 A
re
a
 
Carbon Number 
600°C 
Figure 4.3: Distribution of compounds in condensed pyrolysis oil by number of carbons in
molecule from batch pyrolysis of HDPE. Figures for 450°C and 550°C have been omitted to
conserve space, but show no significant deviations from interpolating between those figures
shown. Summary statistics for all temperatures are presented in table 4.3.
4.5 Analysis of Condensed Pyrolysis Oils and Waxes 85
Table 4.3: Comparison of the condensed pyrolysis products produced with activated carbon and
coke reactor beds. Both carbon number (CN) and molecular weight (MW, g mol-1) are compared,
and the mean, standard deviation (SD) and maximum for the analysed samples are given. The
proportion of condensed products that fall into the liquid transport fuel range (C≤21 — a carbon
number of 21 or less) as a proportion of total peak area is also given. Mean and standard
deviation were calculated using weightings based on peak area percentages.
Activated Carbon Coke
400°Ca 450°C 500°C 550°C 600°C 400°Ca 450°C 500°C 550°C 600°C
CN 10.3 10.5 10.0 10.6 11.5 25.0 33.0 32.7 32.2 29.9
MW 142.0 143.5 137.5 145.3 159.0 351.1 462.4 456.2 450.9 418.4
CN SD 3.7 3.8 4.0 5.2 6.3 14.6 13.5 14.2 14.5 14.5
MW SD 52.9 52.5 56.9 74.2 89.2 204.7 189.7 198.4 203.8 203.4
CN Max 30 30 32 32 35 55 53 55 55 55
MW Max 422.5 422.5 450.5 450.5 492.6 772.9 742.8 772.9 772.9 772.9
C≤21 98.1% 98.3% 97.5% 94.0% 90.8% 51.7% 25.5% 28.9% 30.8% 35.7%
a400°C was the lower threshold at which pyrolysis occurred; the resulting prolonged reactor residence time
increased cracking, though operating batch apparatus at this temperature would not be practical or efficient
in reality, and these figures should be read in this context.
the hydrocarbons produced with an activated carbon bed lie within a much smaller
range (C5 to C35 vs. C5 to C55, also reflected by the lower standard deviations). These
characteristics — lighter products over a smaller range of sizes — account for the visual
differences observed in the products described earlier.
4.5.3.1 Operation at the Temperature Threshold for Pyrolysis
The oil produced at 400°C using a coke reactor is significantly lighter than that
produced at other temperatures. Operating at 400°C using either bed resulted in the
pyrolysis process occurring extremely slowly — at or near the threshold of it occurring
at all, and the process was still not complete even an hour after the HDPE sample was
added. The same lower threshold has been observed by other researchers pyrolysing
HDPE [186, 202], and is consistent with thermogravimetric analyses, which show that
the thermal degradation begins at, or just below this temperature [203]. In the coke
bed case, the low reactor temperature dramatically increased the time the input sample
spent in the reactor; the pyrolysed compounds were heated to a lower temperature,
thus requiring them to be cracked to a greater extent before they entered the gaseous
state and could leave the reactor — accounting for the lighter oil produced at this
temperature. In the activated carbon case, this effect is small in comparison with the
extra cracking from the catalytic effect. In practice, the dramatically longer reaction
time would make operating at this temperature with the batch addition of sample
infeasible.
86 Batch Pyrolysis of HDPE using an Activated Carbon Reactor Bed
4.5.3.2 Influence of Reactor Temperature on Cracking Rate and Residence
Time
In most pyrolysis work, reaction temperature has consistently been found to be the
operating parameter that has had the greatest impact on the make-up of the pyrolysis
products [46, 204, 205]. Increasing the reactor temperature has two primary, yet
opposing effects: it increases the reaction rate of the thermal cracking process, so
more pyrolysis occurs in the same period of time; but it can also decrease the residence
time of the process, decreasing the amount of cracking that occurs.
The input sample undergoes pyrolysis only when it has sufficient energy by being
above the threshold temperature at which thermal breakdown occurs — in microwave-
assisted pyrolysis this effectively means that the sample must be in the “hot zone” of
the reactor bed, where it is conducting heat energy from the carbon.
Sample leaves the reactor bed only when it has been cracked sufficiently that it
enters the gaseous state and physically egresses to the condensing system. A higher
bed temperature means that sample will enter the gaseous state sooner, and may thus
leave the reactor bed with less cracking. This reduced residence time has been shown
to be a large determinant of pyrolysis product composition [46, 206]. Furthermore,
at higher temperatures, the increased rate at which cracking occurs means there
is greater positive pressure generated as solid sample enters the gaseous state and
expands, which provides a second impetus for sample to leave the hot zone of the
reactor bed.
In microwave-assisted pyrolysis the targeted bulk heating of the carbon bed exag-
gerates the contrast in temperatures between the reactor hot zone (it being the only
microwave receptive material in the reactor) and anything present outside this area.
Overall, then, it is the balance of cracking rate and residence time that determines
the composition of the products for a particular temperature. This can vary depending
on reactor configuration: almost all researchers report that the increased cracking
effect dominates, and the average molecular mass decreases with increasing reactor
temperature [173, 204, 205, 207]; however, in the only comparable microwave-assisted
pyrolysis work, Ludlow-Palafox observed an increase in molecular mass of the pyrolysis
condensate when reactor temperature was increased from 500°C to 600°C, which he
attributed to a decrease in sample residence time [29].
4.5 Analysis of Condensed Pyrolysis Oils and Waxes 87
4.5.3.3 Influence of Reactor Temperature on Molecular Mass Distribution of
Products
While it is clear that in this work the use of a catalytic reactor bed resulted in changes
to the pyrolysis products that far outweighed any temperature-induced variation, there
are, nevertheless, temperature-dependent trends within each bed type. At increased
reactor temperatures, shorter residence time and greater rate of cracking act in
concert to increase the range of molecular weights in the products (an increase in
smaller products from the greater cracking, and an increase in larger products from the
higher temperature causing molecules to enter the gaseous state and exit the reactor
sooner). This is observed in both the activated carbon and coke cases, with the range
of molecular masses increasing most at 550–600°C.
Generally, the similarity of the profiles of the compounds produced using an
activated carbon bed across the temperatures studied is remarkable, though the tail of
the distribution begins to lengthen and encompass longer molecules at temperatures
over 550°C, with a corresponding increase in average molecular mass and standard
deviation. For condensed products produced using a coke bed, a similar distribution of
products is observed between 450–550°C, with the average molecular mass decreasing
beyond this point as reactor temperature increases. While this increase in cracking
dominates in the coke bed case, consistent with the research described in §4.5.3.2, the
average molecular mass increases at higher temperatures in the activated carbon case.
Because of the catalytic activity of the activated carbon, the cracking rate is already
high even at lower tempatures; the increase in range of molecular masses facilitated
by operating at a higher reactor temperature (i.e., from causing larger molecules to
enter the gaseous state and leave the reactor, and from reduced residence time due
to faster phase change volume expansion of the cracked HDPE) must then cause the
overall average to increase.
4.5.3.4 Bimodal Molecular Mass Distribution with Coke Reactor Bed
The bimodal distribution of molecular masses in the condensed pyrolysis products
produced using a coke bed suggests the possibility that more than one mechanism
may have resulted in the presence of the heavier modal grouping in the condensate.
Furthermore, this grouping contains within it at least some hydrocarbons with a boiling
point greater than the reactor temperature, and must have therefore arisen through a
different means than vaporizing in the reactor and re-condensing in the condensing
88 Batch Pyrolysis of HDPE using an Activated Carbon Reactor Bed
system. One possible mechanism accounting for the presence of these compounds
is that they are concurrently physically transferred by the continuous flow of gases
exiting the reactor. This may be through physical transfer processes such as chain
entanglement, surface tension where gaseous and longer liquid molecules are held
together (analogous to the fluid present around a soap bubble), or convection; or
alternatively via chemical recombination, with chains re-polymerising, as suggested
by Williams and Williams [173], who also observed a significant fraction of condensed
product that exhibited a boiling point higher than the reactor temperature would have
reached.
That this mechanism does not apparently occur in the activated carbon case could
be due to the smaller molecules produced that do not entangle, the formation of a large
portion of more chemically stable aromatic species, which would not re-polymerise,
or simply the greater bed surface area, leading to greater (simultaneous) cracking of
the HDPE so that the resulting pyrolysis products all fall below the molecular mass
where secondary physical transport becomes an issue. In any case, with or without
the additional heavy molecular grouping it is clear that the condensate produced using
an activated carbon bed is considerably lighter and has undergone substantially more
cracking.
4.5.3.5 Production of Liquid Transport Fuels
The fraction of condensed products that fall within the range of the liquid transport
fuels (C5–C21) particularly emphasises the increased cracking that the input HDPE
undergoes when the activated carbon bed is used. Indeed, as table 4.3 shows, >98%
of the oil produced using an activated carbon bed at temperatures of 450°C and lower
falls within this liquid transport fuel range. With the standard coke bed, an increase in
these useful C≤21 products is observed as increased reaction temperature (and thus
energy expenditure) promotes increased cracking (the 400°C scenario being a special
case discussed in 4.5.3.1); nevertheless, as reaction temperature increases with an
activated carbon bed, the proportion of condensed product that lies within the valuable
liquid transport fuel range actually decreases, as hydrocarbons are produced across
a wider range of masses. In other words, activated carbon produces a narrower and
more useful range of products as reactor temperature is reduced (with a lower bound of
sufficient thermal energy being input for the pyrolysis process to occur in a reasonable
amount of time).
4.5 Analysis of Condensed Pyrolysis Oils and Waxes 89
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
400 450 500 550 600
T
o
ta
l 
P
e
a
k
 A
re
a
 
Pyrolysis temperature (°C) 
Activated Carbon 
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
400 450 500 550 600
T
o
ta
l 
P
e
a
k
 A
re
a
 
Pyrolysis temperature (°C) 
Coke 
n-Alkane Aliphatics n-Alkene Aromatic Cycloalkane
Cycloalkene iso-Alkane iso-Alkene Alkyne
Figure 4.4: Chemical composition of condensed pyrolysis products produced using activated
carbon (top) and coke (bottom) reactor beds with batch introduction of sample. Here, “alkenes”
encompass species containing one or more double bonds. “Aliphatics” includes linear alkanes
and alkenes which were unable to be distinguished from one another in the GC/MS analysis
owing to their large molecular mass (>C23), though dashed line represents best estimate of the
division of this grouping between n-alkanes and n-alkenes based on identified species.
4.5.4 Chemical Composition of Condensed Pyrolysis Products
Figure 4.4 compares the composition of the condensed products from the microwave-
assisted pyrolysis of HDPE using activated carbon and coke beds. As the hydrocarbon
chain length gets longer, it becomes more difficult to distinguish between different
types of aliphatic species with the same carbon number, because the distinguishing
moiety of the molecule (e.g., two carbons connected by a double bond) becomes
an increasingly small proportion of the whole (and correspondingly the similarity
90 Batch Pyrolysis of HDPE using an Activated Carbon Reactor Bed
22 23 24 25 26 27 28 29 30
0.5
   1.0
 1.5
Retention Time (mins)
Ab
un
da
nc
e (
x1
07 )
Figure 4.5: Merging of aliphatic gas chromatograph peaks at larger molecular masses. While
the alkadiene, alkene, and alkane peaks are clearly separate at a retention time of 22 minutes,
they slowly merge and become indistinguishable by the 29–30 minute mark.
between molecules grows, with increased chain length coming in the form of identical
alkyl groups). As shown in figure 4.5, the gas chromatography column becomes
progressively less able to separate these species (as their chemical properties and
retention times converge), until around C23 the peaks merge completely. These species,
comprising linear alkanes and alkenes, have been grouped together in this analysis as
“aliphatics”.
With a standard coke bed the condensed oil is comprised of straight chain alkanes
and alkenes, with around 3% or less of the total comprising aromatic species. This
agrees with the observed consensus in the literature that linear aliphatic hydrocarbons
comprise the bulk of the condensed products resulting from the pyrolysis of HDPE (in
an inert medium), regardless of reactor type.
In FTIR analysis of the condensate produced from the conventional pyrolysis of
polyethylene using a fluidised bed reactor, Williams and Williams [173] noted the
significant formation of aromatic species only at 700°C (these temperatures were
beyond the operating maximum for the rig used in this work).
Also using a fluidised bed, Hernández et al. [208] did not see any aromatic species in
the condensed pyrolysis products at all, with linear aliphatics comprising the entire oil;
a large increase in the alkene to alkane ratio of the pyrolysis condensate from 500°C to
600°C was also observed, consistent with this work, where although the proportion of
alkanes to alkenes is approximately even from 400°C to 500°C, above this temperature
the proportion of alkenes increases.
In the only directly comparable microwave-assisted pyrolysis experimental work,
Ludlow-Palafox and Chase [29] observed similar trends to this work using a coke bed,
4.5 Analysis of Condensed Pyrolysis Oils and Waxes 91
with the alkene:alkane ratio increasing with increasing reactor temperature; however,
the overall proportion of species other than linear aliphatics was higher in their work,
varying from 18.9–7.2% peak area from 500–700°C. The exact make-up of this fraction
is not given, though some of this no doubt comprises branched and cyclic aliphatics.
Nevertheless, the results of this work using a coke bed fall well within Ludlow-Palafox’s
“general statement that the pyrolysis of HDPE yields little aromatics (<5%)” [46].
Each pyrolytic scission of the polyethylene backbone creates two highly reactive rad-
icals that stabilise themselves predominantly through intermolecular or intra-molecular
hydrogen transfer, creating alkanes and alkenes respectively (for a full discussion of
the molecular mechanisms underlying the pyrolytic breakdown of HDPE see §6.2).
Based on McCaffrey et al. [209], who showed that intermolecular hydrogen abstraction
is blocked by an increase in hydrogen abundance, Ludlow-Palafox theorised that
the observed increase in proportion of alkenes with increased reaction temperature
was due to an increase in hydrogen content (though not necessarily in the form of
molecular hydrogen gas) in the pyrolysis gas [46], possibly resulting from an increase in
repolymerisation reactions at higher temperatures [206, 210]. While this link remained
unproven in that research owing to inadequate GC column resolving ability at the low
molecular masses these hydrocarbon gases exhibit, further evidence in partial support
of this theory is presented in §4.6.2.
The catalytic activity of the activated carbon bed is clearly evident in the pyrolysis
oil produced from it. While simple straight chain alkanes and alkenes still make up
a significant fraction of the oil, there is a dramatic increase in proportion of aromatic
compounds (35.5–45.3%), and the presence of other types of hydrocarbons is observed
(cyclo- and branched alkanes and alkenes, as well as alkynes), increasing to 10% peak
area of the oil at 600°C. These figures are in excess of any non-catalytic derived results
reported in the literature at equivalent temperatures.
The increase in alkenes at the expense of alkanes as reaction temperature increases
is also observed using an activated carbon bed, likely as a result of the reasons outlined
above, and it is exactly this reducing environment coupled with a dearth of hydrogen as
the HDPE is progressively cracked that promotes the formation of the more “complex”
hydrocarbons previously mentioned.
The variety of compounds produced by the activated carbon bed relative to the
coke bed is also notable: more than 150 different compounds are present in the
activated carbon derived oil compared with typically around 100 for coke-bed derived
oil, and these compounds occur over a smaller range of carbon lengths. Much of this
92 Batch Pyrolysis of HDPE using an Activated Carbon Reactor Bed
extra variety derives from the greater quantity of aromatic compounds present in the
activated carbon derived oil, though many of these are present only in very small
quantities (the three main aromatic species are toluene, ethylbenzene, and xylene,
with peak areas of 4.8–8.8%, 3.6–5.4%, and 1.6–5.3% respectively; the peak area
percentages decline rapidly after this point). Indeed, in general the pyrolysis products
exhibit a “long tail” distribution, with many different compounds present in small
quantities as a result of the stochastic decomposition process. The aromatic compounds
are primarily single ring benzene derivatives, made up of varying and/or multiple alkyl
groups bonded to a central benzene ring; the presence of these compounds in such
quantities points to activated carbon catalysing the formation of these aromatic rings.
4.6 Analysis of Non-condensable Pyrolysis Gases
4.6.1 Mass Distribution of Non-condensable Pyrolysis Gases
The bulk of the non-condensable pyrolysis gases (table 4.4) are two or three carbons
in length, independent of the type of reactor bed used. As the reactor temperature
increases, more cracking occurs and smaller gases are produced; this is the case when
using both activated carbon and coke beds, though the activated carbon pyrolysis gases
are overall very slightly heavier, with longer chains, than the coke ones at the same
reactor temperature. In spite of this, methane (the only C1 compound) is not produced
in large quantities at any temperature, and hydrogen gas is not produced at all (no ions
of the appropriate mass were found in scans of the chromatogram). As the hydrocarbon
with the highest hydrogen to carbon ratio, the formation of methane is selected against
Table 4.4: Comparison of the non-condensable pyrolysis products produced with activated
carbon and coke reactor beds by carbon number and peak area percentage. Average carbon
number (CN) and molecular weight (MW, g mol-1) are calculated using the weightings derived
from peak area.
Activated Carbon Coke
400°Ca 450°C 500°C 550°C 600°C 400°Ca 450°C 500°C 550°C 600°C
C1 2.8 2.9 3.1 5.3 6.8 4.6 3.6 4.2 4.5 5.8
C2 28.8 27.2 30.9 37.6 40.5 46.1 39.9 42.9 45.6 54.0
C3 55.2 59.0 55.9 49.6 49.0 45.1 49.6 44.1 36.4 31.5
C4 13.3 9.4 9.2 4.8 3.8 4.3 6.9 8.8 13.5 8.6
C5 0.0 1.6 1.0 2.7 0.0 0.0 0.0 0.0 0.0 0.0
CN 2.8 2.8 2.7 2.6 2.5 2.5 2.6 2.6 2.5 2.4
MW 40.8 40.6 39.8 38.0 36.2 35.7 37.3 38.0 36.7 34.5
a400°C was the lower threshold at which pyrolysis occurred; the resulting prolonged reactor residence
time increased cracking, though operating batch apparatus at this temperature would not be practical or
efficient in reality, and these figures should be read in this context.
4.6 Analysis of Non-condensable Pyrolysis Gases 93
in the hydrogen-deficient environment of the reactor. Molecular hydrogen is not
present, owing to the highly reducing carbon bed, and the fact that hydrogen radicals
would quickly bond to the myriad of newly cracked hydrocarbon radicals present (C=C
bond formation is more energetically difficult than C-H). This finding is consistent
with other researchers who produced little to no molecular hydrogen when pyrolysing
polyolefins at the temperatures considered in this study [11, 29, 202, 207], though
Cozzani et al. [206] observed a significant quantity of hydrogen in the gas resulting
from the pyrolysis of polyethylene in a fixed bed (20–30% gas vol. at 500°C and 600°C).
This particular experimental configuration had a very low residence time caused by
the constant high flow of nitrogen gas directly through the (horizontal) reactor; this
separated out pyrolysis products immediately as they were produced, preventing the
recombination of hydrogen-rich species into an overall more stable configuration, and
resulting in a comparatively high char formation (2–10% wt.) from the remaining
hydrogen-deprived species.
Ludlow-Palafox [46] analysed the non-condensable gases produced from pyrolysis
of HDPE at temperatures of 500°C and 600°C, but was unable to extend the analysis
below C3 owing to insufficient column resolving power. Small quantities of compounds
up to C7 were also present in their samples, with the proportion of gases declining
as the carbon number increased, as seen in the present work. The narrower range of
compounds present in this work is likely due to the superior efficiency of the condensing
system.
The pyrolysis gases generated at 600°C using a coke bed were very similar to those
reported in Donaj et al. [207] when pyrolysing a mixture of polyolefins (76% PE, 24%
PP) at 650°C with C2 and C3 molecules making up the bulk of the gases. The higher
proportion of methane generated in this case is attributable to the presence of the
polypropylene, with its methyl group side chain.
4.6.2 Chemical Composition of Non-condensable Pyrolysis Gases
The non-condensable pyrolysis gases are composed entirely of alkanes and alkenes
(table 4.5 on the next page). By virtue of their small molecular size the compounds
are simple — almost all linear, with trace quantities of branched molecules. As
reactor temperature increases a greater proportion of alkenes are produced, with a
corresponding decrease in alkanes, analogous to the change observed in the condensed
pyrolysis products. Given the same absolute number of hydrogen atoms, the increased
cracking at higher temperatures seen in the analysis of molecular masses necessitates
94 Batch Pyrolysis of HDPE using an Activated Carbon Reactor Bed
Table 4.5: Percentage chemical composition of the non-condensable pyrolysis gases produced
with activated carbon and coke reactor beds, with batch introduction of HDPE. Figures given are
peak area percentages. H/C is the overall proportion of hydrogen to carbon calculated for each
gas based on the proportional peak area of each compound.
Activated Carbon Coke
400°Ca 450°C 500°C 550°C 600°C 400°Ca 450°C 500°C 550°C 600°C
n-Alkane 85.0 69.7 67.8 62.7 56.9 40.7 41.5 38.8 33.5 27.8
n-Alkene 15.0 26.3 28.4 32.8 39.6 57.9 51.6 55.5 62.4 66.9
iso-Alkane 0.0 2.4 2.9 1.8 1.6 1.4 2.6 3.2 1.6 1.2
iso-Alkene 0.0 1.6 1.0 2.6 1.9 0.0 4.3 2.5 2.5 4.0
H/C 2.68 2.57 2.57 2.58 2.56 2.41 2.37 2.31 2.28 2.28
a400°C was the lower threshold at which pyrolysis occurred; the resulting prolonged reactor residence
time increased cracking, though operating batch apparatus at this temperature would not be practical or
efficient in reality, and these figures should be read in this context.
the formation of individually less hydrogen-dense molecules (i.e., alkenes instead of
alkanes).
The marked increase in alkanes over alkenes in the non-condensable gases produced
using an activated carbon bed is due to the greater proportion of “complex” unsatur-
ated molecules formed in the corresponding condensate, with the resulting hydrogen-
containing radicals forming fully saturated gases. The higher H/C proportion of the
activated carbon gases compared with the coke gases supports this, with the only other
hydrocarbon pyrolysis product (the condensate) therefore having a correspondingly
lower H/C proportion, resulting from the greater proportion of these more complex
molecules with a lower H/C proportion (aromatics, cycloalkanes, cycloalkenes, alkynes)
described in §4.5.4.
Interestingly, the H/C proportion remains approximately constant across reactor
temperatures for gases produced using an activated carbon bed (the 400°C case
excepted), in spite of the increase in proportion of alkenes. This can be explained by
the greater overall proportion of alkenes at higher temperatures being balanced by the
greater cracking occurring, which results in smaller molecules, with higher individual
hydrogen content.
In contrast, the H/C proportion of the non-condensable gases produced with a coke
bed reduces somewhat with increasing temperature, as there is both less change in
molecule size and a greater proportional increase in alkenes. This observation is
initially apparently in opposition to the earlier discussed relationship Ludlow-Palafox
proposed regarding an increase in hydrogen abundance in the pyrolysis gases causing a
blocking of intermolecular hydrogen abstraction, and the observed decrease in alkanes
and increase in alkenes in the condensate at higher temperatures — certainly, the
proportion of hydrogen in the gases produced using a coke bed falls. However, while
4.7 Conservation of Mass and Hydrogen Transfer 95
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
0 5 10 15 20 25 30 35
H
/C
 
Carbon Number 
Alkanes
Alkenes
Alkadienes
Aromatics
Figure 4.6: Hydrogen to carbon proportion of unbranched hydrocarbons by carbon number.
The aromatics in this case are the most common class observed in the condensate: benzene
derivatives with alkyl groups that increase in length with carbon number.
the relative amount of hydrogen per unit mass is smaller, the absolute amount of
hydrogen-rich gas is in fact much greater, owing to a higher overall amount of non-
condensable gas being produced, as shown in §4.8. Furthermore, it is difficult to
establish a causative relationship in this case, as an increase in the formation of
alkadienes would equally result in an increase in hydrogen-rich radicals and ultimately
gases.
4.7 Conservation of Mass and Hydrogen Transfer
The principle of conservation of mass requires there to be the same total number of
carbon and hydrogen atoms in the unprocessed HDPE as in all the pyrolysis products.
The stoichiometry of polyethylene is such that H/C = 2 in the input product. As figure
4.6 shows, a pyrolytic decomposition of HDPE into pure alkenes would not change the
H/C balance (and would approach a return to the ethene monomers from which HDPE
is produced as cracking continued). Alkanes are more hydrogen-rich than polyethylene,
and alkadienes less, and this imbalance is amplified as the molecules become shorter
and the functional group comprises a larger relative proportion of the molecule as a
whole. Nevertheless, for every hydrogen-rich molecule that is produced, an equally
hydrogen-deficient one must also be, in order that the overall number of atoms is
conserved. Indeed, if, for every alkane produced an equivalent alkadiene (of the
96 Batch Pyrolysis of HDPE using an Activated Carbon Reactor Bed
same carbon number) were also produced the H/C balance would remain satisfied.
Given that these three classes of hydrocarbon are the simplest breakdown products
of HDPE (reaction pathways discussed in §6.2) it is no surprise that they are also the
most common species produced when pyrolysing HDPE through conventional means,
as demonstrated in §4.5.4.
This conservation of mass, and hydrogen transfer, underlies all of the final pyrolysis
product compositions, and the mechanisms by which they are produced. Indeed, the
whole system exists in a dynamic equilibrium where thermal scission, inter- and intra-
molecular hydrogen abstraction, and further secondary reactions interplay to create
the complex array of species observed in the final pyrolysis products.
By comparing the products produced with an activated carbon bed with those
produced using a coke bed, it is clear that the activated carbon promotes secondary
reactions and the creation of aromatics and other complex products. As the aromatics
are hydrogen poor (relative to polyethylene), their formation “frees up” hydrogen
and allows the formation of a greater portion of alkanes in the gaseous state, as
previously observed, implying that a dearth of hydrogen is a limiting factor influencing
the chemical make-up of the gaseous fraction (especially in the coke bed case). This
is further emphasised when it is considered that only alkanes can exist below C2;
alkadienes cannot physically form below C3, nor aromatics below C6, thus as products
are cracked into progressively smaller molecular sizes they are forced to assume more
hydrogen-rich conformations, resulting in the high H/C make-up of the non-condensable
gases previously described.
4.8 Pyrolysis Product Component Yields
Table 4.6 shows the component product yields from the microwave-assisted pyrolysis
of HDPE. Consistent with other work in the microwave pyrolysis field [29, 149], reactor
temperature is the parameter that has the largest single effect on the constituent make-
up of the pyrolysis products. As reaction temperature increases, more cracking occurs:
more gaseous products result, with a corresponding decrease in condensed products.
Given that the gaseous products have a high H/C ratio, as discussed, the greater
cracking and increased proportion of gases at higher reaction temperatures impacts
the make-up of the condensate, necessitating the formation of a greater portion of
lower H/C species, which explains the presence of these species in higher proportions
observed in §4.5.4.
4.8 Pyrolysis Product Component Yields 97
Table 4.6: Percentage yields of retained, condensed, and non-condensable pyrolysis products
with batch introduction of sample. Retained refers to the portion of input mass left in the reactor
at the conclusion of the experiment, condensed product is the oil/wax collected in the condenser,
condensed C≤21 is the subset of the condensed products with a carbon number less than or equal
to 21, non-condensable (gaseous) product accounts for the remaining mass (and is calculated as
such). All percentages are expressed on a mass basis as a proportion of the input HDPE, using
peak area as a determinant of the proportion of products that lie at or below C21.
Activated Carbon Coke
400°Ca 450°C 500°C 550°C 600°C 400°Ca 450°C 500°C 550°C 600°C
Retained 10.3 9.0 6.0 7.3 11.3 2.0 1.0 0.0 0.0 0.0
Condensed 54.9 50.3 41.0 30.7 27.3 59.1 69.0 42.0 42.3 31.4
Cond. C≤21 53.8 49.5 40.0 28.8 24.8 30.6 17.6 12.1 13.0 11.2
Gases 34.8 40.7 53.0 62.0 61.3 38.9 30.0 58.0 57.7 68.6
Overall C≤21 88.6 90.2 93.0 90.8 86.1 69.5 47.6 70.1 70.7 79.8
a400°C was the lower threshold at which pyrolysis occurred; the resulting prolonged reactor residence time
increased cracking, though operating batch apparatus at this temperature would not be practical or efficient
in reality, and these figures should be read in this context.
In spite of the higher percentage of condensible products produced with a coke bed
(reaching a maximum of 69.0% of input mass at 450°C), much of these are less useful
long-chain waxes (as shown in §4.5.3); when only products in the liquid transport fuel
range (C≤21) are considered, using the activated carbon bed results in significantly
higher yields at all temperatures. This is also the case if the same criterion is applied
to all products (both gases and condensibles), with the bulk of the “useful” products
comprising gases produced at high temperatures when using a coke bed, in any case.
In other words, more useful products are produced using an activated carbon reactor
bed than with a coke one, and this is possible at lower temperatures.
At low temperatures using a coke bed the retained fraction represents un-pyrolysed
HDPE. The significant increase in the retained fraction when using an activated carbon
bed required further investigation (see §6.6), though it may be that the activated carbon
catalyses not only hydrocarbon cracking but also the formation of char, which remains
within the reactor.
4.8.1 Comparison with Other Published Work
Table 4.7 on the following page presents a comparison of the yields of the products
resulting from the pyrolysis of polyethylene, in batch operation, between this work
and all other identified comparable peer-reviewed publications. There is considerable
variation in the results reported across the literature, a fact that emphasises the
importance of the reactor type and specific configuration, both of which themselves
vary considerably across the published work. This reinforces the need to perform
control experiments in this work using the same reactor but with a coke bed, in order to
98 Batch Pyrolysis of HDPE using an Activated Carbon Reactor Bed
T
a
b
le
4
.7
:
C
o
m
p
a
riso
n
o
f
P
E
p
yro
lysis
yie
ld
s
in
b
a
tch
o
p
e
ra
tio
n
fro
m
4
0
0
–6
0
0
°C
.
P
e
rce
n
ta
g
e
yie
ld
s
o
f
n
o
n
-co
n
d
e
n
sa
b
le
g
a
se
s,
co
n
d
e
n
se
d
p
ro
d
u
ct,
a
n
d
m
a
ss
re
ta
in
e
d
in
th
e
re
a
cto
r
a
t
th
e
co
n
clu
sio
n
o
f
th
e
e
xp
e
rim
e
n
t,
a
re
a
ll
re
p
o
rte
d
o
n
a
m
a
ss
b
a
sis
a
s
a
p
ro
p
o
rtio
n
o
f
th
e
in
p
u
t
sa
m
p
le
.
N
.R
.
=
n
o
t
re
p
o
rte
d
;
F
B
R
=
fl
u
id
ise
d
b
e
d
re
a
cto
r;
C
S
B
R
=
co
n
ica
l
sp
o
u
te
d
b
e
d
re
a
cto
r;
T
G
A
=
th
e
rm
o
g
ra
vim
e
tric
a
n
a
lysis.
R
e
fe
re
n
ce
R
e
a
cto
r
T
yp
e
R
e
a
cto
r
V
o
lu
m
e
S
a
m
p
le
V
o
lu
m
e
H
e
a
tin
g
M
e
th
o
d
B
e
d
T
yp
e
T
e
m
p
e
ra
tu
re
R
e
ta
in
e
d
C
o
n
d
e
n
se
d
G
a
se
s
T
h
is
W
o
rk
S
tirre
d
B
e
d
1
2
L
1
0
0
g
M
icro
w
a
ve
A
ctiva
te
d
C
a
rb
o
n
4
0
0
°C
1
0
.3
5
4
.9
3
4
.8
4
5
0
°C
9
.0
5
0
.3
4
0
.7
5
0
0
°C
6
.0
4
1
.0
5
3
.0
5
5
0
°C
7
.3
3
0
.7
6
2
.0
6
0
0
°C
1
1
.3
2
7
.3
6
1
.3
T
h
is
W
o
rk
S
tirre
d
B
e
d
1
2
L
1
0
0
g
M
icro
w
a
ve
C
o
k
e
4
0
0
°C
2
.0
5
9
.1
3
8
.9
4
5
0
°C
1
.0
6
9
.0
3
0
.0
5
0
0
°C
0
.0
4
2
.0
5
8
.0
5
5
0
°C
0
.0
4
2
.3
5
7
.7
6
0
0
°C
0
.0
3
1
.4
6
8
.6
A
ch
ilia
s
e
t
a
l.
[2
1
1
]
F
ixe
d
B
e
d
N
.R
.
1
g
E
le
ctric
N
o
n
e
4
5
0
°C
6
1
.0
3
8
.5
0
.5
A
l-S
a
le
m
a
n
d
L
e
ttie
ri
[2
1
2
]
T
G
A
N
.R
.
1
5
m
g
E
le
ctric
N
o
n
e
5
0
0
°C
0
.1
7
9
.7
2
0
.2
5
5
0
°C
0
.2
6
7
.0
3
2
.8
6
0
0
°C
0
.2
5
6
.0
4
3
.7
C
o
zza
n
i
e
t
a
l.
[2
0
6
]
F
ixe
d
B
e
d
1
.2
L
1
5
–2
0
g
E
le
ctric
N
o
n
e
5
0
0
°C
~
2
–5
~
8
3
–9
0
~
8
–1
2
6
0
0
°C
~
6
–1
0
~
5
5
–7
4
~
2
0
–3
5
E
lo
rd
i
e
t
a
l.
[2
0
5
]
C
S
B
R
3
L
N
.R
.
E
le
ctric
S
a
n
d
6
0
0
°C
N
.R
.
~
9
4
~
6
H
e
rn
á
n
d
e
z
e
t
a
l.
[2
0
8
]
F
B
R
1
.9
L
2
g
E
le
ctric
S
a
n
d
5
0
0
°C
3
6
.5
4
8
.3
1
5
.2
6
0
0
°C
2
4
.6
4
3
.0
3
2
.4
K
u
m
a
r
a
n
d
S
in
g
h
[1
8
6
]
S
e
m
i-b
a
tch
1
5
6
m
L
2
0
g
E
le
ctric
N
o
n
e
4
0
0
°C
4
.6
1
1
.2
8
4
.2
4
5
0
°C
3
.8
2
4
.0
7
2
.2
5
0
0
°C
3
.0
7
2
.2
2
4
.8
5
5
0
°C
2
.5
7
9
.1
1
8
.4
L
u
d
lo
w
-P
a
la
fo
x
a
n
d
C
h
a
se
[2
9
]
S
tirre
d
B
e
d
4
L
5
0
g
M
icro
w
a
ve
C
o
k
e
5
0
0
°C
0
.0
8
1
.0
1
9
.0
6
0
0
°C
0
.0
7
9
.1
2
0
.9
M
a
rcilla
e
t
a
l.
[2
1
3
]
C
ru
cib
le
N
.R
.
6
0
0
m
g
E
le
ctric
N
o
n
e
5
5
0
°C
0
.0
8
4
.7
1
6
.3
W
illia
m
s
a
n
d
W
illia
m
s
[1
7
3
]
F
B
R
5
.5
L
3
g
E
le
ctric
S
ilica
S
a
n
d
5
0
0
°C
0
.0
8
9
.2
1
0
.8
5
5
0
°C
0
.0
7
8
.6
2
1
.4
6
0
0
°C
0
.0
7
5
.8
2
4
.2
4.8 Pyrolysis Product Component Yields 99
derive useful comparative results. Very few results were reported below 500°C, likely
due to the longer reaction time when operating below this temperature.
4.8.1.1 Temperature Dependent Gas Production and Influence of Reactor
Configuration
Several trends are evident across the results considered, such as the increase in the
proportion of non-condensable gases produced as reactor temperature and cracking
correspondingly increase. The one exception to this trend is the work presented
by Kumar and Singh [186], the examination of which serves a useful purpose in
demonstrating the importance of reactor configuration. In this study, HDPE was heated
in a reaction vessel with a long vertical exit pipe that reached sufficiently high to allow
room for the condenser to fit in the space below. In this case it is thought that the length
of the pipe, and its lack of heating or insulation, would present a high “barrier to exit”
for pyrolysis products leaving the reactor. As such, they would likely continually reflux
and crack into progressively smaller products until they were sufficiently small that
they could remain in the gaseous state for a sufficient period of time that they could
leave the reactor. Most other experimental configurations used the continual flow of an
inert gas to ensure an oxygen-free environment, but which also served to convectively
force the pyrolysis gases from the reactor. In this case the only impetus for evolved
gases to leave the reactor was the volume expansion pressure generated from cracking
the HDPE into gaseous molecules, and this had to overcome both gravitational and
thermodynamic pressures, potentially increasing the residence time of the reaction as
products were unable to escape the reactor. Indeed, in this extreme case, the reactor
configuration influenced the residence time to the point that it became the dominant
process parameter, evidenced by the extremely long reaction times reported (well over
12 hours at 400°C decreasing to a minimum of 54 minutes at 550°C) — far in excess of
those reported by any other researcher. By contrast, the conical spouted bed reactor
used by Elordi et al. [205] gave a “residence time of the volatiles in the range of centi-
seconds”, and resulted in the highest reported recovery of condensed products of 94%.
4.8.1.2 Minimal Retained Portion
Most experiments showed very little mass retained in the reactor at the end of the
experiment, with the exceptions to this trend apparently deriving from the experiment
not running through to completion for one reason or another. In Achilias et al. [211],
where 61% of the input sample was retained in the reactor at the end of the pyrolysis,
100 Batch Pyrolysis of HDPE using an Activated Carbon Reactor Bed
the methods mention “the time of the experiment was 17 min”, an apparent externally
imposed constraint given that the range of plastics considered in this work would
require different times to pyrolyse. Here, the residue fraction is not discussed further.
In the work of Hernández et al. [208], which saw 25–37% of the input sample
retained in the bed at the conclusion of the experiment, the entirety of the fluidising and
pyrolysis gas “flow [was] shifted to the sampling bag” for collection and later analysis.
Given a reported fluidising gas flow of 3.5 L/min, and a 25 L gas sampling bag, this
constrains the maximum experimental run time to seven minutes, even if no pyrolysis
gas were produced, and this time constraint is likely the source of the high retained
portion.
In this work, the pyrolysis was allowed to run through to completion; the minimal
retained product remaining in the coke bed case is consistent with other researchers.
The higher retained fraction in the activated carbon case is investigated further in §6.6.
4.8.1.3 Relative Gaseous and Condensed Product Yields and the Influence of
Reactor Configuration used in This Work
A maximum condensed product yield of around 70–90% was seen across the examined
body of research; this most commonly occurred at a reactor temperature of 500°C.
In this work, in the comparable coke bed case, a maximum condensed product yield
of 69% at 450°C was observed, which is at the lower end of the range reported in
the literature. Indeed, generally an increase in the yield of gaseous products was
observed in this work for any given temperature; this can be explained when the
specific characteristics of the reactor used to undertake this work are taken into
account.
The reactor used in this work is the largest reactor in this literature comparison, by
volume, by a factor of greater than two. The larger volume reactor creates a greater
intervening space between the hot reactor bed and the condenser system — a space in
which the pyrolysis products may cool, recondense, and undergo secondary reactions,
and a scenario exacerbated by the targeted microwave heating mode used in this work
where only the reactor bed is heated. Furthermore, the reactor in this work does not
operate in a direct “pass-through” mode whereby the products are directly forced out of
the reactor hot zone, unlike many of the compared reactors. Instead, the inert nitrogen
gas necessarily enters from the top of the reactor (see figure 3.14), and may pass
over the reactor bed when pyrolysis is under way. In this way the pyrolysis products
must crack sufficiently to travel up and out of the reactor before condensing. It is
4.8 Pyrolysis Product Component Yields 101
the combination of these factors and the higher barrier to exit that resulted that likely
caused the greater cracking and comparably larger portion of non-condensable gaseous
products.
In addition, in this work, partially as a consequence of operating a larger reactor, the
flow-rate of inert gas in proportion to reactor volume of 0.06 L/min/L, is considerably
less than that of others such as Hernández et al. [208] with 1.84 L/min/L, or 0.25
L/min/L in the more directly comparable work of Ludlow-Palafox and Chase [29]. The
egress of pyrolysis products is therefore likely to be primarily driven by phase change
volume expansion of the products themselves. It should be noted that a full-scale
pyrolysis plant is unlikely to use inert carrier gas at all, though it was used in this case
for safety purposes, and to create an initial oxygen-free environment before pyrolysis
started.
When primary products are forced to egress from the reactor quickly, captured
products provide information about the initial chemistry of the pyrolytic breakdown.
However, many of the reactor configurations required to achieve this state are only
practical on a laboratory scale, and provide little information about the behaviour
of larger systems required for scale-up, and the secondary reactions and subsequent
products that result. Indeed, these secondary reactions have been shown to be critical
when dealing with larger reactors [202, 214], and when optimising for a particular
product [215], and are discussed further in §6.2.
4.8.1.4 Comparison to Other Activated Carbon Work
As the only other work using activated carbon to catalytically crack polyethylene,
the work of González et al. [197] bears some discussion. In this study, polyethylene
supermarket bags were catalytically cracked (with a 1:10 mass ratio of activated
carbon:plastic) using a small electric fixed bed furnace through which flowed inert
nitrogen gas. The highest yield of total products of 49.2% of input mass (corresponding
to 50.8% retained in the reactor) was obtained with activated carbon at 450°C. At
this temperature the products comprised all aliphatic species except for around 6%
aromatics.
The remarkably large proportion of char retained in the reactor (this is not likely
to have been unpyrolysed HDPE as the experiment was run over two hours) is likely
because of a combination of the activated carbon and the flow-through design of this
reactor, where the nitrogen carrier gas results in a very low residence time for the
produced pyrolysis products. This constant flow combined with the small size of the
102 Batch Pyrolysis of HDPE using an Activated Carbon Reactor Bed
reactor means that almost as soon as products entered the gaseous state they would
leave the hot zone and not undergo further cracking. As such, hydrogen-rich alkanes
(formed through activated carbon catalysed hydrogen transfer reactions as described
in §6.3.1) are swept out of the reactor, resulting in the greater production of char as the
remaining input mass has less hydrogen with which to form cracked pyrolysis species.
This is supported by the higher methane content of the evolved gases in this study
(22% vs. 2.9%), and the lower aromatic content (6% vs. 30%), as aromatic and non-
linear species require a greater formation time in the hot zone than linear aliphatics
[178, 179].
The use of microwave heating and a stirred bed as in this dissertation, and the
introduction of sample when the reactor was at the target temperature as opposed
to heating from room temperature, may have also contributed to a more even heat
distribution and less resulting char formation. In any case, the work of González et al.,
like those studies discussed above, may provide information about pyrolysis breakdown
in a specific low-residence configuration, but such configurations are less useful as
models for larger industrial processes.
4.9 Utility of Products Produced Using an Activated
Carbon Bed
4.9.1 Condensed Pyrolysis Oil
The condensed pyrolysis oil produced at lower temperatures using activated carbon
comprises more than 98% hydrocarbons C≤21, and suggests that it would be very
easy to use this microwave-assisted pyrolysis to produce a high value, saleable product
from the pyrolysis oil with a minimum of extra processing. Given the tunability of
diesel engines and the breadth of fuels that they can be run on [216–219], it should be
possible to run a suitably configured engine on the pyrolysis oil directly. The carbon
length profiles suggest that it would be possible to directly separate the pyrolysis oil
into petrol and diesel fractions, though the appropriate standards [220, 221] would also
have to be met before it could be sold on the open market.
Compared with the condensed product produced using the conventional coke bed,
it is clear that the activated carbon bed has, in effect, performed in situ upgrading of
the pyrolysis oil, resulting in a much lighter and potentially more useful liquid product.
Indeed, the heavier aliphatics present in the coke bed derived oil have little value by
4.9 Utility of Products Produced Using an Activated Carbon Bed 103
themselves and would usually be actively detrimental to any product containing them,
though they can be used as a feedstock in crackers to produce other shorter chain
products. All of the produced pyrolysis products could be used as fuels in conventional
fossil fuel boilers, though this does not maximise the potential chemical value that could
be extracted from the plastic.
4.9.1.1 Aromatic Content
The condensed products produced using an activated carbon bed contained a high
proportion of aromatic compounds; for example, the four most common aromatics
present in the oil were, in order, toluene, ethyl-benzene, xylene, and benzene. While
the addition of aromatics to liquid transport fuels increases the octane/cetane rating
and reduces knocking, the maximum aromatic content permissible in petrol and diesel
has been capped in many countries in recent times for environmental reasons [222].
If the pyrolysis oil were to be used as a liquid transport fuel, it is likely that these
components would have to be separated out — by solvents and liquid-liquid extraction,
or extractive distillation. Fortunately, these compounds can have an even greater value
than aliphatics [223], finding use as solvents, and as precursors for a wide variety
of compounds including drugs, lubricants, detergents, plastics (e.g., polystyrene,
polycarbonate), and explosives.
4.9.2 Gases
The non-condensable gases produced during the microwave-assisted pyrolysis of HDPE
are also economically important, the primary constituents being ethane, ethene,
propane, and propene (these making up over 85% peak area of the gas). These high-
energy gases could be used on-site in a generator, in order to power the microwave
heating and make the process energetically self-sustaining. Alternatively, the pyrolysis
gases could be split into the relevant fractions and fed into the existing distribution
network for LPG, natural gas, and natural gas liquids. Ethene and propene in particular
have a multitude of uses, and indeed are the single most used organic chemical
feedstocks worldwide [224], with the creation of virgin HDPE and PP amongst their
uses.
104 Batch Pyrolysis of HDPE using an Activated Carbon Reactor Bed
4.10 Summary
This chapter presents the most complete characterisation of the products resulting
from the microwave-assisted pyrolysis of HDPE to date, and represents the first
characterisation of the microwave-assisted pyrolysis of HDPE using an activated carbon
bed. Large differences were observed between the products resulting from the
different reactor beds, with the condensed pyrolysis oil produced using an activated
carbon bed approximately three times lighter than that produced using a coke bed
(with an average carbon number of 10–11.5 vs. 25.0–33.0); furthermore, the activated
carbon oil contained hydrocarbons with a much narrower range of molecular masses.
The molecular masses of the oil derived from an activated carbon bed matched
those of the liquid transport fuels petrol and diesel (C5–C21), with over 98% of the
condensed product, and 50% of the total input mass converted into these fractions
when operating at a reactor temperature of 450°C, a temperature lower than most
conventional pyrolysis is performed at. Increasing the reactor temperature above this
point increased the cracking and lowered the exit threshold for pyrolysed molecules,
resulting in a product containing an increased average and range of molecular masses.
The pyrolysis products produced using a coke bed were almost all linear aliphatic
compounds, in agreement with other researchers. Indeed, the condensed fraction
contained a high proportion of less useful long-chain waxes. By contrast, using
an activated carbon bed resulted in a large increase in “complex” molecules in the
condensed products, primarily aromatics (35.5–45.3% peak area of the condensate),
which were themselves mostly single ring benzene structures and derivatives (the most
common being toluene, ethyl-benzene, xylene, and benzene itself).
The non-condensable gases were mainly two or three carbons long, independent of
reactor type, and were comprised of simple linear aliphatics by virtue of their small
molecular size. Methane was not present in large amounts owing to the molecule’s
high proportion of hydrogen to carbon (hydrogen being a rate-limiting reactant owing
to the stoichiometry of HDPE); hydrogen gas itself was not observed. The conservation
of mass meant that with an activated carbon bed, the formation of less hydrogen-dense
aromatics in the condensate “freed up” hydrogen, and a greater alkane content was
observed in the activated carbon derived gases.
Overall, it is clear that microwave-assisted pyrolysis of HDPE with an activated
carbon bed produces a narrower range of products that are potentially more valuable
and useful, than when a conventional coke bed is used. The larger reactor configuration
4.10 Summary 105
used in this work promotes secondary reactions, resulting in a greater proportion of
gaseous products, though it is important to understand the influence of these reactions
as they are likely to become more important in any process scale-up. Indeed, the results
presented in this chapter were sufficiently promising to warrant further investigation
into the microwave-assisted pyrolysis of HDPE. Accordingly, as a further step towards
scale-up, a more industrial-like operation was developed where the reactor operated in
a continuous throughput mode and was able to process much larger volumes of sample;
this work is presented in the following chapter.

5
Continuous Pyrolysis of HDPE using
an Activated Carbon Reactor Bed
5.1 Introduction
Having established in the previous chapter that using activated carbon as a catalytic
bed in the microwave-assisted pyrolysis of HDPE was feasible, and having identified
the resulting pyrolysis products as possessing sufficient utility and economic value as
to make the process worthwhile, it was decided to more fully characterise the pyrolysis
process, with the intention of determining whether it would be suitable for scale-up
and operation on a larger, industrial scale. In order to achieve this, it was necessary
to change the configuration of the pyrolysis equipment so that it could process larger
quantities of material, and in a continuous mode with a constant input of HDPE and
collection of products, as the capacity, efficiency, and economic demands of a potential
commercial industrial facility would necessitate.
107
108 Continuous Pyrolysis of HDPE using an Activated Carbon Reactor Bed
Furthermore, while microwave-assisted pyrolysis has been demonstrated in continu-
ous operation when processing input gases [168] and liquids [152], there is a dearth of
reports in the literature dealing with the continuous processing of solid material. This
is likely due to the still burgeoning nature of the microwave-assisted pyrolysis field,
and the inherent difficulties in processing solids in this context — they can quickly
become tacky and difficult to handle when heated; an appropriate feeding mechanism
must allow solid particles to enter the reactor while remaining impervious to returning
pyrolysis gases; to say nothing of the difficulties of designing suitable mechanisms to
operate within the confines of a “microwave friendly” space. Nevertheless, this study
presented an excellent opportunity to investigate the differences between batch and
continuous processing using microwave-assisted pyrolysis, and pioneer the continuous
input of solid material.
Another necessary step towards scaling up the activated carbon catalysed pyrolysis
process was evaluating and quantifying the repeatability of the process and the
products it produced over time, and between independent runs, to determine how
consistent these were, and to investigate the performance of the catalyst over time
— this, in particular, required the ability to process larger masses of HDPE afforded by
the continuous input of sample.
This chapter presents the results and analysis obtained from the continuous pro-
cessing of HDPE with activated carbon catalysed microwave-assisted pyrolysis, com-
paring the process and products to those produced using the batch mode operation
described in the previous chapter, and relates the associated discovered difficulties
resulting from the scale-up of the process to handle larger quantities of input material.
5.2 Method
Continuous feed operation required a redesign of the sample injection apparatus, and
resulted in the continuous rotary feeder described in §3.4.2. HDPE was added to the
reactor at a rate of 11.8 g/min. This speed was selected so that the microwave heating
system could easily maintain the reactor temperature throughout the course of the
experiment, even when operating at the highest studied temperature, thus avoiding
the pitfalls of the batch investigation where at high temperatures the introduction
of sample overwhelmed the microwave heating capability and caused the reactor
temperature to drop. The same addition rate was used across all experiments carried
out. The typical mass of HDPE added throughout an experiment was around 500 g.
5.3 Pyrolysis Product Component Yields 109
The continuous addition study was primarily conducted at 450°C — the reaction
temperature with the greatest yield and without a prolonged residence time — though
a more limited set of comparative data was collected at 400°C, 500°C, and 600°C.
A fresh bed was used when beginning each set of experimental runs at the same
temperature (i.e., those experiments conducted sequentially at the same reactor
temperature). Other than the described changes, the same materials, experimental
apparatus, methodology, and analytical techniques were applied as those described in
the previous chapter (§4.2–4.4).
5.3 Pyrolysis Product Component Yields
5.3.1 Coke Bed Control Yields
The continuous pyrolysis process was first applied to the coke bed in order to establish
consistency data for the control non-catalytic state; three experimental repeats were
performed at each temperature. Table 5.1 shows the average yields obtained at each
temperature. The maximum yield of condensed products of 84.1% was obtained at
450°C. This yield then rapidly dropped off with increasing reaction temperature as the
fraction of input HDPE converted to non-condensable gases increased, until the ratio
of condensed to non-condensable gases reversed at 600°C, with 81.5% of the input
HDPE converted to non-condensable gas. Almost no input sample was retained within
the reactor, and the yield between runs was relatively consistent, with a maximum
standard deviation of 3.5%.
The larger quantity of HDPE processed in these runs, combined with the relatively
heavy products produced when using the coke bed, caused issues in some cases where
the percolating dry ice condenser became blocked. In the event that this occurred, the
experiment was cut short, and the yield percentages were calculated from the quantity
of HDPE that was actually added during the experiment.
Table 5.1: Percentage yields of retained, condensed, and non-condensable pyrolysis products
produced using a coke reactor bed with a continuous input feed of HDPE. Retained refers to
the portion of input mass left in the reactor at the conclusion of each experiment, condensed
product is the oil/wax collected in the condenser, gases refers to the remaining non-condensable
gaseous fraction and is calculated by difference. All percentages are expressed on a mass basis
as a proportion of the input HDPE ± the standard deviation.
400°C 450°C 500°C 600°C
Retained 0.1 ±0.1 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0
Condensed 80.4 ±3.0 84.1 ±3.5 52.5 ±3.2 18.2 ±3.3
Non-condensable gases 19.5 ±3.1 15.9 ±3.5 47.5 ±3.2 81.5 ±3.3
110 Continuous Pyrolysis of HDPE using an Activated Carbon Reactor Bed
As with the batch experiments, greater cracking occurs at higher pyrolysis temper-
atures, resulting in an increased yield of non-condensable gases. At 400°C this trend
is countered by an increase in residence time as the reactor temperature approaches
the lower temperature limit for pyrolysis to occur. However, even at this temperature
with the longer residence time the reactor did not become “overwhelmed” with input
solids undergoing processing. A comparison between the results generated in batch
and continuous input mode is given in §5.7.
5.3.2 Activated Carbon Yields
A similar procedure of attempting to determine the consistency of the microwave-
assisted pyrolysis process was applied using the activated carbon bed; however, it
quickly became clear that large differences in the component yields were occurring
between runs. Figure 5.1 shows the changes in yields with each successive run as
more HDPE was processed with the activated carbon bed. Testing at 500°C and 600°C
was stopped after six experiments each, as the resulting condensed product was highly
viscous and began to clog the condenser.
5.3.2.1 Retained Fraction
The mass fraction of the input HDPE that is retained in the reactor bed is initially very
large (ranging from 15–28% depending on temperature, with a larger fraction retained
as the operating temperature increases), but rapidly reduces as the cumulative mass
of HDPE processed by the bed increases. Indeed, from the extended 450°C case the
retained fraction seems to tend towards an asymptote of zero.
5.3.2.2 Condensed Products
The proportion of condensed pyrolysis products is also temperature dependent, and
follows a similar order to that observed in the coke bed case, with the greatest
yield observed at 450°C, and the yield rapidly declining as the reaction temperature
increases from here. As more sample is processed with the activated carbon bed,
the yield of condensed products increases; in the 450°C case the yield appears to be
approaching that of the coke bed equivalent and, given the similar progression of the
other temperatures for the examined data, it is likely that they too would approach
their coke bed yield equivalents in time.
5.3 Pyrolysis Product Component Yields 111
0%
10%
20%
30%
40%
50%
60%
70%
80%
0 1 2 3 4 5 6 7 8
Y
ie
ld
 (
%
 I
n
p
u
t 
M
a
s
s
) 
HDPE Mass Pyrolysed (Normalised to Bed Mass) 
Retained Fraction 
400°C
450°C
500°C
600°C
0%
10%
20%
30%
40%
50%
60%
70%
80%
0 1 2 3 4 5 6 7 8
Y
ie
ld
 (
%
 I
n
p
u
t 
M
a
s
s
) 
HDPE Mass Pyrolysed (Normalised to Bed Mass) 
Condensed Products 
400°C
450°C
500°C
600°C
0%
10%
20%
30%
40%
50%
60%
70%
80%
0 1 2 3 4 5 6 7 8
Y
ie
ld
 (
%
 I
n
p
u
t 
M
a
s
s
) 
HDPE Mass Pyrolysed (Normalised to Bed Mass) 
Non-condensable Gases 
400°C
450°C
500°C
600°C
Figure 5.1: Percentage yields of retained, condensed, and non-condensable pyrolysis products
produced using an activated carbon reactor bed with a continuous input feed of HDPE. The
retained fraction refers to the portion of input mass left in the reactor at the conclusion of the
experiment; condensed product is the oil/wax collected in the condensing system; the remaining
fraction comprises non-condensable gases and is calculated by difference. All percentages are
expressed on a mass basis as a proportion of the input HDPE.
112 Continuous Pyrolysis of HDPE using an Activated Carbon Reactor Bed
5.3.2.3 Non-condensable Gases
For the most part the yield of the non-condensable gases declines over time as the
mass of HDPE processed increases, as would be expected given the increase in
condensed products. The seeming counter-example to this trend is the initial increase
in proportion of non-condensable gases produced at the 600°C reaction temperature —
this is the result of the comparatively larger decrease in the proportion of sample that
is retained in the reactor over this period: the 18% drop in retained mass represents
additional material that is converted to condensed and gaseous products, such that
the net effect on the non-condensable gases is to increase over this period. If the
experiments at this temperature had been able to be continued and more HDPE was
processed, the non-condensable fraction would likely decrease in line with the results
obtained for the other temperatures, and the final data point gathered shows initial
evidence of this occurring. When the retained fraction is excluded, independent
of the amount of HDPE processed, the non-condensable yields show a temperature
dependence consistent with all the data observed thus far, with the greatest portion of
gases produced at 600°C, and this decreasing as reaction temperature is reduced to
450°C, before slightly increasing at 400°C.
5.3.2.4 Change in Activity of Bed
From these longitudinal studies, conducted with significantly larger quantities of HDPE
than the initial batch input investigation of the previous chapter, it is clear that a
constitutive change in the nature of the pyrolysis products produced using the activated
carbon bed is occurring as more HDPE is processed. The increase over time in the
proportion of heavier condensed products, and decrease in the lighter non-condensable
gases that have undergone greater cracking, is consistent with a reduction in catalytic
activity of the activated carbon bed. This is supported by a visual observation of the
condensed pyrolysis products, which over time became more viscous, and less clear
and homogeneous. That a significant proportion of the input mass remains in the
reactor (the “retained” fraction) even after being subject to temperatures that would
normally drive off any volatiles present, suggests the production of char, where the
catalyst has effectively stripped the hydrocarbon of hydrogen, leaving a carbonaceous
residue. In this way, the retained fraction is effectively hydrocarbon that has been
cracked “beyond” the non-condensable gases stage (and deposited on to the surface of
5.4 Analysis of Condensed Pyrolysis Oils and Waxes 113
the bed particles), explaining the initial increase in these gases (and decline in retained
fraction) at 600°C as catalytic deposition reduces.
This accumulation of mass retained in the reactor over time, the corresponding
apparent loss of catalytic activity of the activated carbon bed, and the decrease in
the rate of accumulation as the activity of the bed itself declines, suggests that the
catalytic deactivation may be occurring via coking, with the deposition of non-active
carbon on the surface of the activated carbon particles. This is supported by the lack of
any significant mass retained in the reactor when operating with the non-catalytic coke
bed. Indeed, by the end of the experimental series conducted at 450°C, the activated
carbon bed had accumulated (non-catalytic) material equivalent to 25% of its mass. If
nothing else, assuming this material is non-catalytic, this accumulation of mass would
“dilute” the reactor bed and reduce the proportion of catalyst that the HDPE substrate
is exposed to. A detailed investigation into the specific nature of the retained fraction
was later conducted and is presented in §6.6. A general trend from these results,
though, is as more mass is processed by the activated carbon bed, yields tend more
closely towards those obtained when using a coke bed — consistent with the catalytic
deactivation of activated carbon, and with the coke bed representing the ultimate non-
active state.
5.4 Analysis of Condensed Pyrolysis Oils and Waxes
5.4.1 Mass Distribution of Condensed Pyrolysis Products
A complete GC/MS analysis of the condensed products produced in the continuous
mode was carried out, and from this the distribution of molecular masses for each
experiment was obtained. Figure 5.2 on the next page shows the average molecular
mass, and the range of the molecular masses of the compounds present in the condens-
ate (delineated by 5th and 95th percentile so as to eliminate the non-representative
influence of any small concentration outliers).
5.4.1.1 Coke Bed
The distribution of condensed products produced using a coke bed with a continuous
input of sample follow a similar pattern to those produced with batch input. Both
the average molecular mass and the range of masses present in the condensate
increases with increasing reactor temperature, up to 550°C, at which point the average
114 Continuous Pyrolysis of HDPE using an Activated Carbon Reactor Bed
0 1
0
2
0
3
0
4
0
5
0
0
1
0
0
2
0
0
3
0
0
4
0
0
5
0
0
6
0
0
7
0
0
0
1
2
3
4
5
6
7
8
Carbon Number 
Molecular Mass (g mol-1) 
H
D
P
E
 M
a
s
s
 P
y
ro
ly
s
e
d
 (N
o
rm
a
lis
e
d
 to
 B
e
d
 M
a
s
s
) 
A
c
tiv
a
te
d
 C
a
rb
o
n
 
4
0
0
°C
4
5
0
°C
5
0
0
°C
6
0
0
°C
0 1
0
2
0
3
0
4
0
5
0
0
1
0
0
2
0
0
3
0
0
4
0
0
5
0
0
6
0
0
7
0
0
4
0
0
°C
4
5
0
°C
5
0
0
°C
6
0
0
°C
Carbon Number 
Molecular Mass (g mol-1) 
T
e
m
p
e
ra
tu
re
 
C
o
k
e
 
F
ig
u
re
5
.2
:
A
ve
ra
g
e
m
o
le
cu
la
r
m
a
ss
o
f
co
n
d
e
n
se
d
p
yro
lysis
p
ro
d
u
cts
w
ith
co
n
tin
u
o
u
s
in
tro
d
u
ctio
n
o
f
H
D
P
E
.
C
a
p
p
e
d
ce
n
tra
l
e
rro
r
b
a
rs
sh
o
w
sta
n
d
a
rd
d
e
via
tio
n
o
f
m
e
a
n
.
N
o
n
-ca
p
p
e
d
lo
n
g
e
r
b
a
rs
sh
o
w
5
th
a
n
d
9
5
th
p
e
rce
n
tile
o
f
th
e
ra
n
g
e
o
f
th
e
m
o
le
cu
la
r
m
a
sse
s.
A
ll
re
su
lts
a
re
ca
lcu
la
te
d
o
n
p
e
a
k
a
re
a
b
a
sis.
C
a
rb
o
n
n
u
m
b
e
r
a
xis
is
b
a
se
d
o
n
th
e
a
ve
ra
g
e
co
rre
la
tio
n
to
m
o
le
cu
la
r
m
a
ss.
5.4 Analysis of Condensed Pyrolysis Oils and Waxes 115
molecular mass reduces slightly. The average molecular mass of the condensed
products is determined by the amount of cracking these products undergo, which
itself is determined by temperature and residence time. Up to 550°C the increase
in temperature reduces the amount of time the HDPE spends in the hot zone of the
reactor so it is overall cracked less, and has a higher average molecular mass. At
600°C the increase in the speed of the cracking process dominates and the average
molecular mass is reduced. The range of products increases with reaction temperature
as the hotter temperatures allow larger compounds to enter the gaseous state and
exit the reactor, and the more rapid cracking (and thus production of pyrolysis gas
which drives exit from the reactor through phase change volume expansion) reduces
residence time.
While this residence time based explanation of the increase in average molecular
mass of condensed products between 400–500°C certainly holds true, it is important to
remember that this average only encompasses the condensed fraction of the pyrolysis
products produced, and that while the condensed products may have a greater average
molecular mass at 500°C than 450°C, overall, more cracking occurs at 500°C (and
indeed overall cracking consistently increases with reactor temperature), as is clear
from the increased yield of shorter chain non-condensable gases seen in figure 5.1.
5.4.1.2 Activated Carbon Bed
When the activated carbon bed was used, while the temperature and residence time
effects described for the coke bed case are no doubt present, the composition of the
condensed product is primarily determined by the catalytic nature of the activated
carbon. It is clear that over time, as more HDPE is processed by a given activated
carbon bed, both the average molecular mass and the range of masses of the condensed
pyrolysis products increase, consistent with a reduction in bed catalytic activity. The
rate of this increase is highly temperature dependent: at 450°C the initial increase in
average mass of the condensed pyrolysis products is almost negligible, remaining at
around 110–120 g mol-1 (corresponding to around C8–C9) until around 3.5 times the
mass of the bed itself has been processed, after which it increases to a maximum of
300 g mol-1 (around C14), and at which time the rate of increase plateaus. The results
obtained at 400°C follow a similar trend to the 450°C case. However, at 500°C the
average molecular mass increases far more rapidly as more HDPE is processed, with no
initial plateau as in the lower temperature cases; the range of the molecular masses of
the products produced at this temperature increases equally quickly. These increases
116 Continuous Pyrolysis of HDPE using an Activated Carbon Reactor Bed
occur even more rapidly at 600°C, and are consistent with the visual observation of
the products produced at these temperatures, which showed inhomogeneity and high
viscosity (from the greater range and mass of hydrocarbons respectively), and caused
problems with blocking the percolating dry ice condenser.
5.4.1.3 Temperature Dependent Change in Activated Carbon Bed Activity
It is clear that the molecular mass time profile of the products with an activated carbon
bed varies considerably depending on the temperature at which pyrolysis is conducted.
The rapid change in average molecular mass and range of products shown at 500°C
and 600°C, contrasted with the equivalent seen at 400°C and 450°C, implies that the
process leading to the deactivation of the activated carbon bed is highly non-linear with
temperature.
Based on the extended study made at 450°C, as the deactivation of the activated
carbon progresses the profile of the resulting condensed products grows increasingly
similar to those produced using the non-catalytic coke bed. In the 500°C and
600°C cases the composition of the condensed products is still undergoing change;
interestingly, from the third set of experiments (HDPE/Bed Mass ≈ 0.75) the average
molecular mass of the products produced at 600°C is considerably greater than those
produced at 500°C, in contrast with their equivalents produced using the coke bed,
where the average molecular mass is somewhat smaller at 600°C. This may be a result
of the catalytic cracking effectively magnifying the non-linear reduction in residence
time that, with increased reaction temperature, increased the average molecular mass
of the condensed products in the batch mode (§4.5.3.2). If, in the non-catalysed
case, the increase in rate of production of pyrolysis gases at higher temperatures was
sufficient to increase the pressure within the reactor to cause said pyrolysis gases to
leave more quickly (resulting in the higher average molecular mass of the condensed
products), then in the catalysed case the reaction rate will be quicker still, the exit
pressure even greater, the residence time less, and the average molecular mass of the
products greater. Note that it is not the ultimate yield of non-condensable gases that
determines the exit pressure and residence time, but rather the rate of production
of the pyrolysis gas (which includes all products in the gaseous state in the reactor).
Hence, it would be entirely possible for the changes that occurred after those recorded
at 600°C to include a lower average molecular mass in the condensate, and a higher
total proportion of non-condensable gases, as a result of a reduction in catalytic activity
causing an overall slower reaction rate and reducing the residence time.
5.4 Analysis of Condensed Pyrolysis Oils and Waxes 117
5.4.1.4 Change in Carbon Number Distribution
The initial condensed pyrolysis products produced using an activated carbon bed
with continuous addition of HDPE all have very similar molecular mass distributions,
regardless of the temperature at which they were produced, though the upper bound of
the range does increase somewhat with increasing temperature. They are uni-modal,
with the largest peak at C8; the distribution produced at 450°C is shown in figure 5.3
on the following page.
As more HDPE is processed by the activated carbon bed, for all temperatures,
the distribution of the molecular masses changes considerably as the tail grows to
encompass larger molecules. An examination of the extended study conducted at a
reactor temperature of 450°C is useful as this case not only presents the most data, but
the changes in mass distribution happen sufficiently slowly that the snapshots of each
experimental run allow a complete picture of changes to be built up. Data gathered
at 500°C and 600°C is consistent with similar changes taking place but much more
rapidly; at 400°C the change is similar but slower.
At 450°C as more HDPE is processed by the activated carbon bed the distribution
remains similar for some time, until there is a rapid increase in the maximum molecular
mass in the condensed pyrolysis products when the quantity of HDPE processed is
equal to around 3.5 times the mass of the bed: going from 2.9 times the mass of
the bed processed to 4.1 times, the maximum carbon number increases from C20
to C45, accounting for the rapid rate of increase in average molecular mass in this
period observed in figure 5.2. These heavier compounds are initially present in small
quantities, but as more HDPE is processed by the bed they are produced in increasing
quantities, until a clear second grouping becomes evident.
This emergent second grouping resembles that observed in the condensate pro-
duced with the coke bed in the batch mode, is also comprised of long-chain aliphatic
species, and likely derives from the same origin, possibilities of which are discussed
in §4.5.3.4. The increase in prominence of these heavier compounds results in a
corresponding reduction in the lighter grouping and an increasingly flat distribution.
By the time 6.9 times the mass of the bed has been processed, the carbon number
distribution is clearly bi-modal, and highly resembles that produced using the coke bed
at the same temperature, as seen in the lower two graphs of figure 5.3.
118 Continuous Pyrolysis of HDPE using an Activated Carbon Reactor Bed
0%
2%
4%
6%
8%
10%
12%
14%
16%
18%
20%
5 10 15 20 25 30 35 40 45 50 55
%
 T
o
ta
l 
P
ea
k
 A
re
a
 
Carbon Number 
HDPE/AC Bed Mass = 2.7 
0%
2%
4%
6%
8%
10%
12%
14%
16%
18%
20%
5 10 15 20 25 30 35 40 45 50 55
%
 T
o
ta
l 
P
ea
k
 A
re
a
 
Carbon Number 
HDPE/AC Bed Mass = 6.9 
0%
2%
4%
6%
8%
10%
12%
14%
16%
18%
20%
5 10 15 20 25 30 35 40 45 50 55
%
 T
o
ta
l 
P
ea
k
 A
re
a
 
Carbon Number 
Coke Bed 
Figure 5.3: Change in carbon number distribution of condensed products due to deactivation
of activated carbon. The condensed products analysed here were all produced at a reactor
temperature of 450°C with the continuous addition of sample. The top figure represents the
initial mass distribution produced using activated carbon having processed up to 2.7 times its
mass of HDPE; by the end of the experimental series after the bed has processed 6.9 times
its mass of HDPE the mass distribution has changed considerably (middle), and resembles that
resulting from the control coke bed (bottom).
5.4 Analysis of Condensed Pyrolysis Oils and Waxes 119
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
400°C 450°C 500°C 600°C
%
 T
o
ta
l 
P
e
a
k
 A
re
a
 
Pyrolysis Temperature 
Alkyne
iso-Alkene
iso-Alkane
Cycloalkene
Cycloalkane
Aromatic
n-Alkene
Aliphatic
n-Alkane
Figure 5.4: Chemical composition of the condensed pyrolysis products produced using a
coke bed with a continuous input of HDPE. “Alkenes” encompasses species containing one or
more double bonds. “Aliphatics” includes linear alkanes and alkenes which were unable to
be distinguished from one another in the GC/MS analysis owing to their large molecular mass
(>C23).
5.4.2 Chemical Composition of Condensed Pyrolysis Products
5.4.2.1 Coke Bed
Figure 5.4 shows the chemical composition of the condensed pyrolysis products
produced using a coke bed. The bulk of the products comprise linear alkanes and
alkenes (over 90% peak area at 400°C and 450°C), with small quantities of cyclic
compounds. As the reaction temperature increases there is a reduction in the linear
alkanes and alkenes, and an increase in cyclic and aromatic species, which make up
around 25% of the peak area at 600°C.
The proportion of heavier aliphatics (comprising linear alkanes and alkenes, though
indistinguishable from each other in this analysis owing to their large molecular mass)
increases with reaction temperature, as does the ratio of (separable) alkenes to alkanes.
The greater thermal energy at 500°C and 600°C also facilitates secondary reaction
mechanisms (described in §6.2) that lead to the production of aromatics.
120 Continuous Pyrolysis of HDPE using an Activated Carbon Reactor Bed
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0 1 2 3 4 5 6
%
 T
o
ta
l 
P
e
a
k
 A
re
a
 
HDPE Mass Pyrolysed (Normalised to Bed Mass) 
n-Alkane Aliphatic n-Alkene Aromatic Cycloalkane Cycloalkene iso-Alkane iso-Alkene Alkyne
Figure 5.5: Change in chemical composition of the condensed pyrolysis products produced
using an activated carbon bed at 450°C with a continuous input of HDPE. “Alkenes” encompasses
species containing one or more double bonds. “Aliphatics” includes linear alkanes and alkenes
which were unable to be distinguished from one another in the GC/MS analysis owing to their
large molecular mass (>C23–25).
5.4.2.2 Activated Carbon Bed
The change in the chemical composition of the condensed pyrolysis products produced
using an activated carbon bed at 450°C, as it processes progressively more HDPE,
is shown in figure 5.5. The most dramatic change across the range tested is in
the proportion of aromatics, which initially comprise the majority of the condensed
product (around 85% peak area), but as the activated carbon bed processes more HDPE
decrease steadily. The decrease in aromatic content is matched by an increase in linear
alkane and alkene species. After the bed has processed around 3 times its mass in
HDPE we see the appearance of the longer chain indistiguishable aliphatics, which
coincides with the increase in average molecular mass seen in §5.4.1.2. The remaining
classes of compounds (branched aliphatics, cyclics, and alkynes) are never present in
any large quantity, remaining well under 10% peak area at all times even when summed
together.
Figure 5.6 shows the effect of reaction temperature on the change in chemical
composition of the condensed pyrolysis products produced using an activated carbon
bed. The depicted range of masses processed by the beds has been limited in the
450°C case for comparative purposes; however, over this range it is clear that very
5.4 Analysis of Condensed Pyrolysis Oils and Waxes 121
0%
20%
40%
60%
80%
100%
0 1
4
5
0
°C
 
0%
20%
40%
60%
80%
100%
0 1
4
0
0
°C
 
0%
20%
40%
60%
80%
100%
0 1
5
0
0
°C
 
0%
20%
40%
60%
80%
100%
0 1
6
0
0
°C
 
n-Alkane Aliphatic n-Alkene Aromatic Cycloalkane
Cycloalkene iso-Alkane iso-Alkene Alkyne
Y
ie
ld
 (
%
 I
n
p
u
t 
M
a
s
s
) 
HDPE Mass Pyrolysed (Normalised to Bed Mass) 
Figure 5.6: Change in chemical composition of the condensed pyrolysis products produced
using an activated carbon bed with a continuous input of HDPE at all studied temperatures. The
450°C case has been foreshortened in this figure for comparative purposes, but is given in full
in figure 5.5. “Alkenes” encompasses species containing one or more double bonds. “Aliphatics”
includes linear alkanes and alkenes which were unable to be distinguished from one another in
the GC/MS analysis owing to their large molecular mass (>C23–25).
122 Continuous Pyrolysis of HDPE using an Activated Carbon Reactor Bed
little change occurs at reactor temperatures of 400°C and 450°C. In contrast, with
an increase in reactor temperature of just 50°C, the change in species present in
the condensed product over the course of the bed processing its own mass in HDPE
is significant: heavier aliphatic species are immediately apparent, and aliphatics
generally increase over time as the aromatic content declines. In other words, the
same changes observed with the 450°C bed occur, but much more rapidly, consistent
with the activated carbon deactivating more quickly at higher temperatures.
Across the entire range of temperatures and reactor beds, linear aliphatics and
aromatics comprise the vast majority of the pyrolysis products (>90% peak area),
with the ratio of aromatic to aliphatic declining depending on reactor temperature
and, with an activated carbon bed, how much HDPE has been processed. The primary
breakdown products of the pure thermal scission of HDPE are linear aliphatics [225].
Given that aromatics are only present in any significant quantity at high temperatures
(600°C) in the non-catalytic coke bed case, and that they decline as the activated carbon
bed processes more HDPE and becomes deactivated, it is clear that activated carbon
catalyses the formation of aromatic species. Potential reaction mechanisms accounting
for this, and the formation of all other major species observed in the pyrolysis products
are discussed in detail in §6.2.
5.4.2.3 Specific Compound Analysis and Examples
Thus far the chemical analysis of pyrolysis products has focused on general categories
of compounds; tables 5.2 and 5.3 present lists of the most five most common individual
species within each of these categories produced using the coke and activated carbon
beds respectively, with the latter of these also detailing how the proportion of species
changes as the activated carbon progressively processes more HDPE. The stochastic
nature of the pyrolytic breakdown of HDPE results in an extremely large number
of separate compounds produced, making listing all of them impossible given any
reasonable space constraint, and which is not especially instructive in any case given
the category-based analysis already presented. In spite of this, the most common
compounds make up a disproportionately large percentage of the total: the top 24
species across all analyses performed with an activated carbon bed make up 50% of
the total peak area, while the remaining 50% consists of around 550 other species
(though not all of these are present in each analysis). As such, the data presented in
these tables on the most common compounds observed does add informative value, and
provides a representative flavour of the compounds present in each pertinent category.
5.4 Analysis of Condensed Pyrolysis Oils and Waxes 123
Table 5.2: Five most prevalent compounds for each chemical grouping (averaged across all
experiments) in the condensate produced using a coke bed continuously fed with HDPE. All
figures are percent peak area derived from GC/MS analysis. Positional isomers (e.g. o-, m-,
p-xylene) are grouped. Aliphatics are long-chain linear alkanes and alkenes that were not
distinguishable from each other owing to their length; as a grouped category these present a
proportionally larger fraction than the equivalent single species would.
Pyrolysis Temperature 400°C 450°C 500°C 600°C
Alkane
Eicosane 1.5 0.9 0.3 0.2
Docosane 1.5 0.9 0.3 0.1
Octadecane 1.4 0.8 0.3 0.2
Hexadecane 1.3 0.8 0.3 0.2
Heneicosane 1.4 0.9 0.3 0.1
Alkene
Decene 2.6 2.6 2.3 1.6
Undecene 2.2 2.2 1.7 1.3
Heptene 1.5 1.5 1.9 1.1
Tetradecene 2.3 2.0 1.4 1.1
Hexene 1.4 1.4 1.6 1.4
Aliphatic
C36 1.9 2.2 4.1 3.7
C37 1.6 2.1 4.0 3.8
C38 0.8 1.9 3.9 3.9
C35 1.2 2.1 3.9 3.2
C39 0.5 1.8 3.4 3.5
Aromatic
Toluene 0.1 0.4 1.2 4.6
Benzene 0.0 0.2 0.7 4.4
Naphthalene 0.0 0.0 0.1 2.0
Ethylbenzene 0.0 0.1 0.3 1.0
Benzene, 1,3-dimethyl- 0.0 0.1 0.5 0.7
Other
Cyclopropane, ethyl- 0.7 0.8 1.0 0.8
Cyclohexene 0.1 0.2 0.5 0.4
Cyclopentene 0.1 0.2 0.4 0.4
1,3-Cyclopentadiene 0.0 0.1 0.4 0.8
1,3-Cyclohexadiene 0.0 0.1 0.6 0.0
The most prevalent alkanes produced using a coke bed are dominated by the
distribution of compounds at 400°C, which has a higher proportion of alkanes than
the other temperatures. As a result of being so close to the pyrolysis threshold, the
distribution of compounds is also different at this temperature — while the average and
upper molecular mass is smaller than at higher temperatures, the mode is larger as a
result of the flatter distribtuion, accounting for the larger alkanes listed in this table.
Aromatics are not present in large quantities until 600°C, with benzene and derivates
124 Continuous Pyrolysis of HDPE using an Activated Carbon Reactor Bed
T
a
b
le
5
.3
:
F
ive
m
o
st
p
re
va
le
n
t
co
m
p
o
u
n
d
s
fo
r
e
a
ch
ch
e
m
ica
l
g
ro
u
p
in
g
(a
ve
ra
g
e
d
a
cro
ss
a
ll
e
xp
e
rim
e
n
ts)
in
th
e
co
n
d
e
n
sa
te
p
ro
d
u
ce
d
u
sin
g
a
n
a
ctiva
te
d
ca
rb
o
n
b
e
d
co
n
tin
u
o
u
sly
fe
d
w
ith
H
D
P
E
.
A
ll
fi
g
u
re
s
a
re
p
e
rce
n
t
p
e
a
k
a
re
a
d
e
rive
d
fro
m
G
C
/M
S
a
n
a
lysis.
P
o
sitio
n
a
l
iso
m
e
rs
(e
.g
.
o
-,
m
-,
p
-xyle
n
e
)
a
re
g
ro
u
p
e
d
.
S
e
le
cte
d
re
p
re
se
n
ta
tive
ru
n
s
a
re
p
re
se
n
te
d
a
s
th
e
b
e
d
s
h
a
ve
p
ro
ce
sse
d
m
o
re
H
D
P
E
(H
D
P
E
/B
e
d
M
a
ss
=
m
a
ss
fra
ctio
n
o
f
H
D
P
E
to
th
a
t
o
f
th
e
b
e
d
).
A
lip
h
a
tics
a
re
lo
n
g
-ch
a
in
lin
e
a
r
a
lk
a
n
e
s
a
n
d
a
lk
e
n
e
s
th
a
t
w
e
re
n
o
t
d
istin
g
u
ish
a
b
le
fro
m
e
a
ch
o
th
e
r
o
w
in
g
to
th
e
ir
le
n
g
th
;
a
s
a
g
ro
u
p
e
d
ca
te
g
o
ry
th
e
se
p
re
se
n
t
a
p
ro
p
o
rtio
n
a
lly
la
rg
e
r
fra
ctio
n
th
a
n
th
e
e
q
u
iva
le
n
t
sin
g
le
sp
e
cie
s
w
o
u
ld
.
P
yro
lysis
T
e
m
p
e
ra
tu
re
4
0
0
°C
4
5
0
°C
5
0
0
°C
6
0
0
°C
H
D
P
E
/B
e
d
M
a
ss
0
.4
5
1
.1
8
1
.4
2
1
.9
1
0
.5
0
0
.9
9
1
.9
5
3
.3
6
4
.7
8
6
.9
1
0
.4
8
0
.9
7
1
.4
3
0
.4
7
0
.8
6
1
.3
3
A
lk
a
n
e
H
e
xa
n
e
6
.4
5
.6
5
.2
4
.8
5
.2
4
.4
4
.9
2
.4
1
.1
0
.9
2
.7
2
.4
1
.9
2
.1
1
.4
1
.0
P
e
n
ta
n
e
6
.3
5
.1
4
.7
4
.2
4
.8
3
.4
5
.4
1
.7
1
.1
1
.1
3
.9
3
.0
2
.8
2
.1
1
.3
0
.9
H
e
p
ta
n
e
4
.9
4
.8
4
.6
4
.7
3
.9
4
.0
3
.6
3
.0
1
.3
1
.0
2
.5
2
.0
1
.5
1
.4
0
.9
0
.4
O
cta
n
e
3
.5
3
.9
3
.8
4
.3
2
.5
2
.4
2
.5
2
.4
1
.8
1
.1
1
.9
1
.5
1
.3
1
.1
0
.7
0
.3
N
o
n
a
n
e
2
.2
2
.9
2
.9
3
.4
1
.5
2
.4
1
.8
3
.2
2
.2
1
.1
1
.5
1
.0
1
.1
0
.9
0
.6
0
.3
A
lk
e
n
e
H
e
xe
n
e
2
.0
2
.0
1
.9
1
.6
0
.5
1
.1
1
.3
2
.9
1
.3
4
.0
4
.7
3
.9
6
.2
9
.1
5
.5
7
.2
H
e
p
te
n
e
1
.6
1
.8
1
.6
1
.5
1
.6
2
.4
3
.0
1
.2
0
.9
0
.6
2
.5
1
.9
2
.9
3
.5
1
.6
2
.1
O
cte
n
e
1
.1
1
.6
1
.5
1
.5
1
.2
2
.2
2
.1
0
.4
1
.3
1
.4
2
.2
2
.2
2
.1
2
.9
1
.3
1
.9
D
e
ce
n
e
0
.3
0
.3
0
.4
0
.8
1
.0
0
.6
0
.7
2
.1
1
.0
0
.6
1
.1
1
.7
1
.5
2
.2
1
.4
1
.4
U
n
d
e
ce
n
e
0
.2
0
.5
0
.4
0
.4
0
.4
0
.6
0
.5
2
.3
1
.4
1
.8
1
.0
1
.3
1
.3
1
.7
1
.6
1
.8
A
lip
h
a
tic
C
4
0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
1
.0
0
.7
2
.3
1
.3
1
.6
2
.0
0
.2
2
.0
2
.4
C
3
9
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.9
0
.6
2
.2
1
.4
1
.5
1
.9
0
.2
2
.4
2
.5
C
3
8
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.7
0
.5
1
.9
1
.5
1
.3
1
.8
0
.2
2
.8
2
.6
C
4
1
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
1
.0
0
.7
2
.2
1
.2
1
.6
2
.0
0
.3
1
.5
2
.1
C
3
7
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.6
0
.5
1
.5
1
.5
1
.1
1
.5
0
.2
3
.2
2
.3
A
ro
m
a
tic
T
o
lu
e
n
e
1
3
.1
7
.7
7
.5
5
.5
1
3
.6
1
1
.2
9
.2
3
.0
1
.6
0
.9
1
4
.7
5
.7
3
.6
6
.3
3
.2
3
.1
X
yle
n
e
6
.6
4
.9
5
.1
4
.1
9
.9
9
.7
8
.1
5
.4
2
.2
0
.2
4
.5
3
.6
2
.4
1
.8
1
.7
1
.1
E
th
ylb
e
n
ze
n
e
7
.6
5
.3
5
.4
3
.8
9
.9
8
.6
7
.3
2
.8
1
.7
0
.6
4
.1
3
.3
2
.2
2
.3
1
.3
0
.7
B
e
n
ze
n
e
3
.9
1
.7
1
.6
1
.0
3
.2
2
.1
4
.8
1
.4
0
.6
0
.5
4
.7
3
.3
2
.0
4
.4
2
.4
2
.6
N
a
p
h
th
a
le
n
e
4
.5
0
.4
0
.0
0
.0
7
.1
5
.8
6
.3
2
.3
1
.5
0
.2
3
.6
2
.8
1
.9
2
.7
1
.6
0
.9
O
th
e
r
C
yclo
p
ro
p
a
n
e
,
e
th
yl-
0
.8
0
.7
0
.7
0
.6
0
.4
0
.6
1
.0
0
.0
0
.0
0
.0
1
.1
0
.0
0
.0
0
.0
1
.0
1
.5
C
yclo
p
e
n
ta
n
e
,
m
e
th
yl-
1
.1
0
.9
0
.9
0
.8
0
.1
0
.0
0
.0
0
.7
0
.2
0
.2
0
.2
0
.4
0
.3
0
.4
0
.1
0
.0
C
yclo
p
ro
p
a
n
e
,
d
im
e
th
yl-
1
.5
1
.3
1
.2
1
.0
0
.0
0
.0
0
.3
0
.0
0
.6
0
.4
0
.4
0
.0
0
.0
0
.2
0
.0
0
.0
C
yclo
p
e
n
ta
n
e
,
e
th
yl-
0
.4
0
.5
0
.5
0
.5
0
.3
0
.4
0
.3
0
.5
0
.2
0
.1
0
.1
0
.1
0
.1
0
.1
0
.0
0
.0
C
yclo
h
e
xa
n
e
,
m
e
th
yl-
0
.0
0
.4
0
.4
0
.4
0
.2
0
.3
0
.0
0
.7
0
.3
0
.2
0
.2
0
.3
0
.2
0
.3
0
.2
0
.0
5.4 Analysis of Condensed Pyrolysis Oils and Waxes 125
most prevalent, though bi-cyclic napthalene is also present at 600°C. Aliphatics make
up a large proportion of the condensate, though have larger absolute values as this is
a category that groups alkanes, alkenes, alkadienes, etc.
In the condensate produced using fresh activated carbon, toluene is the single most
common species, and remains the most common aromatic species across almost all
testing performed. By far the majority of all the aromatic species are single-ring
benzene derivatives with aliphatic side chains, though benzene itself is also present.
PAHs are present though not in large quantities; naphthalene is by far the most
abundant of these, with the three-ring anthracene the largest PAH observed across
all samples, and this only in small amounts. The quantity of PAHs present declines
quickly as the activated carbon bed processes more HDPE. The most common alkanes
and alkenes are linear, from C5–C11, though as more HDPE is processed by the bed
heavier aliphatics become more common, with a concurrent proportional reduction in
the lighter alkanes and alkenes. Cycloalkanes comprise the majority of the remaining
species, though these are not present in large quantities.
Generally as species contain an increasing number of side chains, or as the side
chains become increasingly branched, they are less prevalent: in order for these
more complex species to form they must undergo a greater number of reaction steps
from the linear primary decomposition products of HDPE. As the number of secondary
reaction steps in a sequence grows, the potential number of possible end-products
grows extremely rapidly, with the probability of a particular individual end-product
forming falling concurrently. Further, the number of secondary reactions a molecule
is likely to undergo is itself probabilistically determined, with a greater number of
reactions in sequence progressively less likely to occur. This combination of facts
accounts for the large number of individual species observed across all the experiments
conducted, and the overall distribution of species, with the smaller numerical subset of
those that occur most frequently making up by far the greatest portion of the whole.
5.4.3 Economic Value and Consistency Over Time of Condensed
Products
From the analysis presented thus far, it is clear that the composition of the condensed
products varies significantly as the activated carbon bed processes progressively more
HDPE. While in the initial studies presented in chapter 4 the condensate showed
significant promise for use in liquid transport fuels, it is necessary to re-evaluate this
in light of the new information on the deactivation of activated carbon and consequent
126 Continuous Pyrolysis of HDPE using an Activated Carbon Reactor Bed
0%
20%
40%
60%
80%
100%
0 1 2 3 4 5 6 7 8
P
e
a
k
 A
re
a
 C
≤
2
1
 
HDPE Mass Pyrolysed (Normalised to Bed Mass) 
400°C
450°C
500°C
600°C
Figure 5.7: Proportion of condensate produced using an activated carbon bed with a carbon
number in the liquid transport fuel range (C≤21). Percentages are derived from peak areas in
GC/MS analysis of the samples.
change in composition of the products presented in this chapter. Figure 5.7 shows
the proportion of the condensed product, produced at various reaction temperatures
with an activated carbon bed, that lies in the liquid transport fuel range (C5–C21).
The rapid deactivation that occurs at operating temperatures of 500°C and greater
results in a swift drop-off of liquid transport fuel range products, making operating
at these temperatures infeasible. At 450°C the entire condensate is initially in the
liquid transport fuel range, and remains so until the bed has processed approximately
three times its mass. As shown in the previous section, over this period the resulting
condensate is highly consistent in chemical make-up and would be ideal for separation
into two valuable output streams: the aromatic BTX fraction, and the aliphatic liquid
transport fuels, as described in the previous chapter. A similar, if slightly slower,
deactivation and processing ability would be expected at 400°C given the relative
similarity observed between these two temperatures. That 450°C is the optimal
temperature for the liquid transport fuel production is reinforced when the yield of
condensed product is considered, which is greatest at this temperature.
Over the long term, however, the value of the catalytic microwave-assisted pyrolysis
process is tempered by the deactivation of the activated carbon and its inability to
produce a consistent output product. In the 450°C case, after the activated carbon
has processed approximately three times its mass of HDPE, heavier aliphatics begin to
5.5 Analysis of Non-condensable Pyrolysis Gases 127
appear in the condensed product. These would require separating out from potential
liquid transport fuel end products, increasing the expense of the overall process. The
balance of desirable shorter chain molecules to heavier products steadily worsens
as the activated carbon processes more HDPE. If the activated carbon bed required
replacement every time after processing only three times its weight of HDPE, the
process would be uneconomical: given the expense of activated carbon, its replacement
cost alone would far outweigh the value of the products produced throughout its
lifetime. A potential solution to this problem would be to maintain catalytic activity
by regularly regenerating the activated carbon. This could potentially be undertaken
in-situ in the reactor itself, a possibility explored in §7.2.4.
5.5 Analysis of Non-condensable Pyrolysis Gases
5.5.1 Chemical Composition
5.5.1.1 Coke Bed
The compounds comprising the non-condensable gases produced using the non-
catalytic coke bed in continuous mode are presented in table 5.4 on the next page.
Alkenes are the most common class of compounds, in line with the non-catalytic
breakdown of HDPE, these being the closest to the monomer from which HDPE is
formed. Compared with the activated carbon case (below), alkanes, with a higher H/C
ratio, are less likely to be present owing to the formation of far fewer catalytically
derived “hydrogen freeing” aromatic species in the condensate.
5.5.1.2 Activated Carbon Bed
Table 5.5 on page 129 presents the results of the GC/MS analysis of the non-
condensable gaseous products produced using the activated carbon bed with a con-
tinuous input of HDPE. The gases are mostly linear alkanes and alkenes, the bulk of
which have a carbon number between C2–C4. The primary factors determining the
make-up of the non-condensable gases are the reaction temperature and the activity
of the activated carbon, which influences cracking and, by altering the make-up of the
larger condensed molecules into more or less hydrogen-dense forms, determines the
quantity of atomic hydrogen “available” for the shorter chain gases as described in
§4.6.2.
As the activated carbon deactivates, there is an increase in the average molecular
128 Continuous Pyrolysis of HDPE using an Activated Carbon Reactor Bed
Table 5.4: Chemical composition of the non-condensable gases produced using a coke bed using
a continuous input of HDPE. Figures are percent peak area derived from GC/MS analysis.
Pyrolysis Temperature 400°C 450°C 500°C 600°C
C1 2.1 2.6 3.9 5.2
Methane 2.2 2.6 3.9 5.2
C2 25.1 28.4 38.8 38.5
Ethane 7.9 8.4 9.0 7.1
Ethene 17.1 20.0 29.8 31.4
C3 65.6 62.5 49.2 44.6
Propane 23.0 18.8 7.6 1.5
1-Propene 42.7 43.6 41.6 43.0
C4 6.7 6.1 8.1 11.4
Butane 2.3 1.7 0.6 0.3
1-Butene 4.1 4.0 5.5 5.0
1-Propene, 2-methyl- 0.3 0.0 0.2 0.9
2-Butene 0.0 0.1 0.5 1.2
1,3-Butadiene 0.0 0.3 1.3 3.9
mass of the non-condensable gases, with a decrease in the smaller gases (e.g, methane,
ethane) and an increase in heavier C4–C5 gases. The proportion of hydrogen-rich
alkanes also decreases as the activated carbon processes more HDPE, mirroring the
decline of aromatic species in the condensed products, from which the hydrogen was
freed that allowed the formation of gaseous alkanes in the first place. This shift is
particularly evident in the shorter chain molecules of the non-condensable gases, as the
hydrogen to carbon ratio necessary for the formation of alkanes increases drastically
as the chain length reduces. Overall, as the activated carbon bed processes more
HDPE the non-condensable gases produced show a steady convergence towards those
produced using a coke bed (in terms of alkene:alkane ratio, average molecular mass,
and relative proportion of species). C4 compounds are more prevalent in the activated
carbon produced gases, the primary component of which is butane; however, it may
be that these compounds decline to the levels seen in the coke produced bed as the
activated carbon becomes more inert — there is evidence of a decline in the sum total
of C4 compounds in the last captured data point at all operating temperatures.
Generally as the reaction temperature increases, greater cracking results in the
gases having a lighter average molecular mass, though this becomes more evident as
the activated carbon deactivates and pure thermal cracking becomes dominant over
catalytic effects.
5.5 Analysis of Non-condensable Pyrolysis Gases 129
T
a
b
le
5
.5
:
C
o
m
p
o
si
ti
o
n
o
f
n
o
n
-c
o
n
d
e
n
sa
b
le
g
a
se
s
p
ro
d
u
ce
d
u
si
n
g
a
n
a
ct
iv
a
te
d
ca
rb
o
n
b
e
d
u
si
n
g
a
co
n
ti
n
u
o
u
s
in
p
u
t
o
f
H
D
P
E
.
A
ll
fi
g
u
re
s
a
re
p
e
rc
e
n
t
p
e
a
k
a
re
a
d
e
ri
ve
d
fr
o
m
G
C
/M
S
a
n
a
ly
si
s.
R
e
p
re
se
n
ta
ti
ve
se
le
ct
e
d
ru
n
s
a
re
p
re
se
n
te
d
fo
r
co
m
p
a
ra
ti
ve
p
u
rp
o
se
s
a
s
th
e
b
e
d
s
h
a
ve
p
ro
ce
ss
e
d
m
o
re
H
D
P
E
(H
D
P
E
/B
e
d
M
a
ss
=
m
a
ss
fr
a
ct
io
n
o
f
H
D
P
E
to
th
a
t
o
f
th
e
b
e
d
).
P
yr
o
ly
si
s
T
e
m
p
e
ra
tu
re
4
0
0
°C
4
5
0
°C
5
0
0
°C
6
0
0
°C
H
D
P
E
/B
e
d
M
a
ss
0
.4
5
1
.1
8
1
.4
2
1
.9
1
0
.5
0
0
.9
9
1
.9
5
3
.3
6
4
.7
8
6
.9
1
0
.4
9
0
.9
7
1
.4
3
0
.4
7
0
.8
6
1
.3
3
C
1
5
.5
4
.1
4
.1
4
.7
5
.6
5
.7
4
.9
3
.1
3
.2
2
.1
1
2
.9
8
.4
6
.0
1
3
.0
9
.1
6
.4
M
e
th
a
n
e
5
.5
4
.1
4
.1
4
.7
5
.6
5
.7
4
.9
3
.1
3
.2
2
.1
1
2
.9
8
.4
6
.0
1
3
.0
9
.1
6
.4
C
2
1
9
.8
1
7
.7
1
1
.3
1
0
.8
3
2
.1
3
5
.0
2
5
.8
2
0
.3
1
8
.9
1
9
.1
3
3
.5
2
0
.9
2
0
.6
2
6
.1
2
4
.1
2
7
.9
E
th
a
n
e
1
6
.3
1
5
.3
8
.9
7
.7
2
8
.1
2
8
.9
2
2
.2
1
6
.1
1
3
.8
8
.9
2
4
.8
1
4
.2
1
2
.0
9
.3
1
0
.9
8
.8
E
th
e
n
e
3
.5
2
.4
2
.3
3
.1
4
.0
6
.1
3
.6
4
.2
5
.2
1
0
.2
8
.7
6
.8
8
.5
1
6
.8
1
3
.3
1
9
.1
C
3
5
3
.8
5
1
.2
4
8
.1
5
6
.9
5
1
.3
5
2
.5
5
3
.3
5
0
.6
5
0
.6
5
8
.1
4
2
.0
4
4
.6
4
8
.8
4
1
.2
4
3
.3
4
2
.0
P
ro
p
a
n
e
4
3
.1
4
1
.9
3
9
.0
4
5
.4
4
0
.4
3
6
.9
3
8
.7
3
1
.6
2
9
.7
2
3
.8
1
8
.8
2
1
.3
2
4
.0
9
.8
1
0
.4
4
.9
1
-P
ro
p
e
n
e
1
0
.7
9
.3
9
.1
1
1
.5
1
0
.9
1
5
.5
1
4
.6
1
9
.0
2
0
.9
3
4
.2
2
3
.2
2
3
.2
2
4
.8
3
1
.4
3
2
.9
3
7
.1
C
4
2
0
.9
2
7
.0
3
6
.1
2
6
.8
9
.3
6
.0
1
4
.8
2
3
.4
2
4
.1
1
6
.8
1
1
.6
2
4
.6
2
0
.3
1
7
.7
2
1
.7
1
9
.3
B
u
ta
n
e
1
7
.1
2
3
.9
2
9
.4
2
0
.9
7
.0
4
.6
9
.8
1
4
.5
1
4
.3
8
.2
6
.7
9
.7
8
.4
3
.1
3
.6
1
.3
1
-B
u
te
n
e
1
.4
3
.1
4
.0
2
.9
0
.4
0
.4
2
.1
3
.4
7
.7
4
.4
4
.9
7
.0
6
.0
6
.5
8
.3
9
.3
1
-P
ro
p
e
n
e
,
2
-m
e
th
yl
0
.0
0
.0
0
.0
0
.9
1
.4
0
.9
2
.9
5
.5
2
.7
4
.2
0
.0
6
.3
1
.9
0
.0
0
.0
1
.9
2
-B
u
te
n
e
2
.4
0
.0
2
.7
1
.4
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
2
.9
3
.5
4
.8
1
.3
1
,3
-B
u
ta
d
ie
n
e
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.5
0
.0
0
.6
0
.0
0
.0
0
.0
3
.9
4
.9
5
.6
P
ro
p
a
n
e
,
2
-m
e
th
yl
-
0
.0
0
.0
0
.0
0
.7
0
.5
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.7
0
.0
0
.0
C
yc
lo
b
u
te
n
e
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
1
.5
1
.2
0
.0
0
.0
0
.0
C
5
0
.0
0
.0
0
.5
0
.8
1
.7
0
.8
1
.2
2
.1
2
.6
3
.3
0
.0
1
.4
4
.3
2
.0
1
.8
4
.4
P
e
n
ta
n
e
0
.0
0
.0
0
.5
0
.8
1
.3
0
.8
1
.2
1
.6
2
.6
1
.5
0
.0
1
.4
3
.1
2
.0
1
.8
3
.6
1
-P
e
n
te
n
e
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.5
0
.0
1
.2
0
.0
0
.0
1
.2
0
.0
0
.0
0
.8
C
yc
lo
p
ro
p
a
n
e
,
1
,2
-d
im
e
th
yl
-
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.5
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
2
-P
e
n
te
n
e
0
.0
0
.0
0
.0
0
.0
0
.4
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
130 Continuous Pyrolysis of HDPE using an Activated Carbon Reactor Bed
5.5.2 Economic Value of Non-Condensable Gases
The non-condensable gases produced with the continuous pyrolysis of HDPE consist
of simple alkanes and alkenes, with similar chemical composition to those produced in
batch mode, and as such retain the same economic potential in terms of use as chemical
feedstock (including for the synthesis of virgin HDPE) and combustion fuels (see
§4.9.2). Indeed, the non-condensable gases by definition encompass a much smaller
range of molecular masses and compounds, and furthermore comprise a smaller subset
of types of hydrocarbon (aromatics, for example, cannot be present owing to the
molecules containing too few carbon atoms to form these rings). There is therefore less
variation in the make-up of the non-condensable gases in spite of potentially significant
underlying changes in bed activity. What variation does exist would not present any
barriers to their utilisation (unlike, for example, the longer chain liquid hydrocarbons
that result when the activated carbon deactivates).
Given the added expense of separating out higher mass aliphatics and aromatics
from any potential liquid fuel product (produced using a coke bed or activated carbon
bed after it has partially deactivated), the natural (and thus free) phase separation
ability of molecules in the gaseous state, the greater uniformity of their chemical
composition, and the still substantial economic value of these resulting gases, it may
be possible to alter the production focus of the pyrolysis process and optimise it such
that hydrocarbon gases are the primary targeted product. This possibility is discussed
in depth as a potential course of further study in §7.2.6.2.
5.6 Temperature Dependence of Rate of Change of
Catalytic Activity
From the analysis of the pyrolysis products presented in this chapter thus far, it is
clear that activated carbon deactivates relatively quickly when used in the microwave-
assisted pyrolysis of HDPE, but further, that this deactivation process itself is highly
temperature dependent: the change in yields and composition of the pyrolysis products
at 500°C and 600°C is far more rapid than that which occurs at 400°C and 450°C.
Assuming the deactivation derives from coking of the pores and/or active sites of
the activated carbon, it may be that the dehydrogenation or equivalent process that
drives this coke formation has an activation energy that causes it to rapidly increase
in rate when this energy threshold is reached. This would explain the large change
5.7 Batch and Continuous Operation Comparison 131
in deactivation rate over the relatively small temperature change between 450°C
and 500°C. A full investigation into the nature of the molecular changes in the
activated carbon accompanying the change in output pyrolysis products, the catalytic
consequences of these changes, and the effect of reaction temperature on the rate of
these changes is given in the following chapter.
5.7 Batch and Continuous Operation Comparison
The development of the continuous rotary feeder allowed the investigation of microwave-
assisted pyrolysis operating in continuous throughput mode as outlined thus far in this
chapter. The results of the previous chapter, using the same procedure and apparatus
operating in batch mode, present an opportunity to compare the characteristics and
product differences between the two modes of operation. Given that the catalytic
nature of the activated carbon bed dramatically impacts on the outputs, and varies
depending on its state of deactivation, the bulk of this comparison focuses on data
generated using the coke bed, though where relevant conclusions can be drawn, the
impact of the differences in addition mode when using activated carbon is also detailed.
5.7.1 Coke Bed Yields
Significant differences in yields were observed between pyrolysis using batch injection
and a continuous feed of HDPE. When operating with a continuous feed, the maximum
yield of condensed products (84.1% at 450°C) is approximately 15% higher than when
operating in batch mode (69.0% at 450°C). A similar deviation was observed in the
maximum non-condensable gases yield (81.5% vs. 68.6% continuous and at batch
respectively at 600°C). At low temperatures relatively more cracking of the HDPE
occurred in the batch mode compared to continuous, but at high temperatures the
reverse was true, and more cracking occurred in the continuous mode. Overall then,
the difference in yields attributable to reaction temperature was considerably larger
when HDPE was added in a continuous fashion.
5.7.2 Chemical Composition of Products Produced using Coke
Bed
The chemical composition of the condensed products is broadly similar between
batch and continuous modes, consisting almost entirely of linear alkanes and alkenes.
132 Continuous Pyrolysis of HDPE using an Activated Carbon Reactor Bed
Table 5.6: Comparison between condensed product properties produced in batch and
continuous reactor modes with a coke reactor bed. Yields of condensed products and non-
condensable gases are given, along with the average molecular mass (MW) and standard
deviation (SD) of the condensed products and overall molecular mass of all combined pyrolysis
products as an indicator of overall cracking.
Batch Continuous
400°C 450°C 500°C 600°C 400°C 450°C 500°C 600°C
Condensed (% wt.) 59.1 69.0 42.0 31.4 80.4 84.1 52.5 18.2
Non-condensable Gases (% wt.) 38.9 30.0 58.0 68.6 19.5 15.9 47.5 81.5
Overall Average MW (g mol-1) 221.4 330.2 213.6 155.0 228.4 258.9 193.5 89.1
Condensed Product MW (g mol-1) 351.1 462.4 456.2 418.4 274.6 300.6 335.4 325.7
Condensed Product SD (g mol-1) 204.7 189.7 198.4 203.4 127.1 141.4 182.2 199.0
However, at reactor temperatures of 500°C and over, the condensate produced in the
continuous mode diverges, with aromatic and cyclic species becoming prevalent (14%
and 25% at 500°C and 600°C in continuous mode vs. 2–4% throughout the batch mode).
5.7.2.1 Molecular Mass Distribution and Heavy Aliphatic Grouping
The average molecular mass of all combined pyrolysis products (i.e., both condensed
and non-condensable gases) is the best indicator of how much cracking of the input
HDPE has occurred over the course of an experimental run. As seen in table 5.6,
which compares batch and continuous modes, it is clear that at temperatures of
450°C and above the pyrolysis products produced in batch mode are significantly
heavier on average (and thus have undergone less cracking) than those produced in
continuous mode. Furthermore, the spread of the pyrolysis products’ molecular masses
is significantly greater in the batch regime for any given reaction temperature. A
significant proportion of this difference derives from the grouping of heavier aliphatic
compounds, which in batch mode are both more numerous, and longer in chain length
(this group having a mode of around C45 in batch configuration compared with C36–C40
in the continuous equivalent).
5.7.3 Residence Time Differences due to Pyrolysis Rate and Exit
Pressure
The fact that the pyrolysis products produced with a continuous input of HDPE have
undergone more cracking than those produced with a batch input can be explained
by considering the pyrolysis rates and residence times of the respective sample
introduction modes. The egress of pyrolysis products from the reactor is primarily
driven by phase change volume expansion of the products themselves as they change
from solid to gas. With the bulk introduction of sample at higher reactor temperatures,
5.7 Batch and Continuous Operation Comparison 133
an initial sudden “explosive” rush of pyrolysis gases was observed entering the
condenser soon after the HDPE was added to the reactor, which then tailed off over
time. Here, the large quantity of introduced material is rapidly pyrolysed and converted
to gaseous form, which expands and provides a large driving pressure differential
forcing gases out of the reactor (“exit pressure”). As the mass of material present in the
reactor declines, the absolute rate of conversion to the gaseous state declines, and so
does the driving force causing pyrolysed material to leave the reactor. This effect was
reduced when operating at 400°C in batch mode, while with the continuous addition of
HDPE, the rate of pyrolysis gas entering the condenser was relatively constant, and no
initial high-velocity “explosive exit” occurred.
5.7.3.1 Simplified Mathematical Model of Input Modes
These observations are perhaps best explained using a simple mathematical model of
the kinetics of the different addition scenarios. Assuming the pyrolysis of HDPE follows
first-order kinetics, i.e., the rate of pyrolysis is proportional to the amount of sample
in the reactor (an approximation used by other researchers [46, 226]), then in batch
mode the pyrolysis process can be modelled by:
dH
dt
= −k(T ).H
where H is the amount of HDPE in the reactor, t is time, and k is the reaction
rate constant (itself primarily dependent on reaction temperature (T ), and often
empirically modelled using the Arrhenius equation [212, 227]). Separating variables
and integrating gives:
H = H0e
−kt
Thus an initial batch injection of HDPE gives an exponentially declining pyrolysis
rate, and an equivalently declining exit force and residence time. Assuming the same
kinetics, but with the continuous addition of HDPE sample gives:
dH
dt
= −k(T ).H +A
where A is the rate at which sample is added to the reactor. By examining the steady
state configuration it is clear that the pyrolysis rate is limited by (and at steady state
equal to) the rate of sample addition.
134 Continuous Pyrolysis of HDPE using an Activated Carbon Reactor Bed
R
a
te
 o
f 
P
y
ro
ly
s
is
 
Time 
Batch Low Temperature
Batch High Temperature
Continuous Low Temperature
Continuous High Temperature
Figure 5.8: Examples of theoretically derived pyrolysis rates in batch and continuous modes at
high and low reaction temperatures.
Figure 5.8 uses the above equations and compares the resulting overall pyrolysis
rate under demonstrative high and low temperature configurations between a single
initial batch introduction of sample and continuous addition scenarios. With the
continuous addition of sample, independent of temperature, the pyrolysis rate quickly
stabilises to the rate of HDPE addition. It is worth noting that even though the pyrolysis
rate stabilises to the same value regardless of temperature, the residence time will
still have a temperature-dependent component, with a longer residence time at lower
temperatures (and a greater mass in the reactor at any given time).
With batch addition, in the high temperature case (demonstrating the scenario likely
occurring when the reactor operates at 450°C and above), the larger reaction rate
constant, and the large initial mass of HDPE present, results in a high initial pyrolysis
rate, which would in turn result in a short residence time for the period in which the
majority of the HDPE is pyrolysed, as large quantities of gas are produced and forced
out of the reactor by volume expansion. In this scenario the low residence time results
in less cracking compared with the continuous addition of sample, where there is a
lower average rate of pyrolysis, longer residence time, and ultimately greater cracking
of the HDPE.
In the low temperature case (corresponding to the 400°C, where operating temper-
ature approaches the threshold of pyrolysis), with a smaller corresponding value for
the reaction rate constant, the pyrolysis rate is much slower, and undergoes slower
decay, potentially remaining below that of the continuously fed case for much of the
experimental run. As such, in spite of the larger amount of sample present in the
5.7 Batch and Continuous Operation Comparison 135
reactor in the batch case, the production of pyrolysis gases is still relatively slow,
and they create less exit pressure to force themselves out of the reactor, and thus
have a longer residence time. So with a slow reaction rate at lower temperatures,
the batch mode results in a longer residence time and greater cracking, with a greater
yield of non-condensable gases and less condensed products compared with continuous
addition.
Not only does this account for the observed differences in cracking, but a second
consequence of the different HDPE addition modes is that while the reaction rate and
residence time quickly stabilise in the continuous addition case, they are very variable
over the course of an experimental run in the batch mode, with the exponential decay
resulting in a long slow rate tail, in contrast to the initial high reaction rate upon sample
addition. The distribution of the molecular masses for the respective products resulting
from these two addition modes should therefore show significant differences at the
same operating temperature, and indeed this is the case. As seen in table 5.6, at any
given reaction temperature the range of molecular masses of the condensed products
produced using a coke bed is considerably larger in batch mode compared with those
produced in the continuous mode.
5.7.3.2 Effect of Residence Time on Chemical Composition of Pyrolysis
Products
Differences in residence time are also able to account for some of the observed
differences in chemical composition between the two reactor modes. In the continuous
mode at reactor temperatures of 500°C and above, significant quantities of aromatic
species were observed in the condensate; these are only present in very small amounts
in the condensate produced in batch mode. The formation of aromatics in the pyrolysis
of hydrocarbons is dependent on both temperature and time: the pyrolysis of decane
by Cypres [228] did not yield any aromatic species below 700°C with a 2 s contact
time, but with increasing contact time the minimum formation temperature decreased
and quantity of aromatic species increased. The described short residence time in
batch mode for the majority of HDPE processed, would, at the operating temperatures
investigated, explain the lack of aromatic species observed in the resulting condensed
products. In the tail of the exponential decay of the pyrolysis rate, as the amount of
HDPE in the reactor declines and the residence time increases, the little hydrocarbon
left may account for the small quantity of aromatics observed, though as the residence
time increases, produced species are more likely to be non-condensable gases in any
136 Continuous Pyrolysis of HDPE using an Activated Carbon Reactor Bed
case due to greater cracking. The longer average residence time resulting from the
continuous mode would also allow for the formation of the observed aromatics at higher
temperatures.
5.7.3.3 Convective Transfer of Pyrolysis Aerosol Particles Owing to Exit
Pressure
The exit pressure generated by the pyrolysis gases not only affects the residence time
in the reactor, but also acts in part to determine which products are able to exit the
reactor in the first place. At least some convective transfer of molecules from the
reactor to the condenser must occur; this is evident from the presence of hydrocarbons
up to C55, which have boiling points of around 600°C, in the condensate, even when the
reactor was operating at a temperature of 400°C. For an aerosol particle (i.e., a small
drop of liquid or solid) to leave the reactor, a sufficiently large force must act upon it
so that it can overcome gravity and escape. This force comes in the form of convection,
where the particle has directional force imparted on it from the collective movement
of the gaseous particles in which it is suspended towards the reactor exit. The greater
the movement of gases (i.e., pressure derived from the pyrolysis induced phase change
volume expansion of the HDPE), the greater convective force will be exerted, allowing
particles with greater mass to overcome gravity and escape the reactor. Given the
pyrolysis rate is much greater as the bulk of the HDPE is cracked into pyrolysis gas in
batch mode, with the large introduction of HDPE to the reactor the convective transfer
will also be greater during this period, with larger heavy-mass molecules able to be
transferred, and in greater quantities. This is borne out by an examination of the mass
distributions of the products produced in the different reaction modes: the grouping
of heavier aliphatic compounds are both more numerous, and longer in chain length in
batch mode compared with the continuous equivalent, and as a result the condensed
products in batch mode are considerably heavier (see table 5.6).
5.7.4 Differences in Reaction Mode for Activated Carbon
Studies
5.7.4.1 Deactivation Across Batch Experimental Set
For the continuous mode experiments detailed in this chapter, a fresh bed of activated
carbon was used at the start of each set of experiments for every reaction temperature
studied. For the initial investigation in the batch mode, before it was clear that the
5.7 Batch and Continuous Operation Comparison 137
activated carbon deactivated relatively rapidly, the same bed was used across all the
experiments (though the bed was fresh at the beginning of these). In spite of this, the
relatively small mass pyrolysed in these experiments compared to the continuous case
should mean that the change in catalytic activity across this set of experiments was
relatively small.
A quick, highly qualitative evaluation of the processed masses involved supports
this argument: each experiment in the batch mode involved 100 g of carbon; the first
experiments at 400°C and 450°C would have seen little change in catalytic activity
due to the slow rate of deactivation at these temperatures. The subsequent three
experiments at 500°C, 550°C, and 600°C involved processing a total of 300 g of HDPE,
a quantity approximately equal to the first data point gathered at each temperature in
the continuous processing case. The results presented in this chapter (e.g. figure 5.1)
indicate little change from baseline with this quantity of HDPE processed. The order of
experiments (lowest temperature to highest) also helps to minimise the total deviation
in catalytic activity from the virgin state. A quantitative theoretical exploration and
validation of this concept is given in §6.5.4.1.
5.7.4.2 Yield and Composition Differences
Comparing the pyrolysis products produced using the activated carbon bed in the batch
mode with the first set of data at each temperature in the continuous mode reveals
similar trends to those observed in the non-catalytic coke bed case. More cracking
of the HDPE occurs in the continuous mode (with a greater yield of non-condensable
gases, and a lighter condensed fraction), with a greater proportion of aromatic species
present in the pyrolysis products. This is consistent with the faster pyrolysis rate
and shorter residence time for the bulk of the HDPE pyrolysed in the batch mode as
described above.
The reduced contact time resulting from the batch mode would also lower the
extra catalytic cracking that occurs with activated carbon, accounting for the smaller
quantity of aromatics, and mass (char) retained in the reactor in the batch case. This
factor is also potentially exacerbated by the better mixing between HDPE and catalyst
that would likely result from the slower continuous addition of sample. That the
observed differences between the two sample addition modes result from the kinetics
underlying these reaction modes, and not the deactivation of the activated carbon over
the course of the batch runs, is supported by the fact that these differences were still
138 Continuous Pyrolysis of HDPE using an Activated Carbon Reactor Bed
observed in the 400°C and 450°C cases, where the activated carbon would be in, or
very close to, virgin condition.
5.8 Summary
The development of the continuous rotary feeder allowed the investigation of the
microwave-assisted pyrolysis of HDPE in continuous throughput mode, facilitating the
processing of an arbitrarily large quantity of sample — a necessary requirement in any
realistic industrial scale-up of the process. This facility was then used to investigate
the variability of the pyrolysis products produced over multiple experimental runs. For
each temperature investigated, using the coke bed resulted in relatively consistent
yields of pyrolysis products, with unchanging chemical composition. By contrast, with
the activated carbon bed, dramatic changes in the pyrolysis products were observed
over time, with the overall cracking reducing (resulting in less non-condensable gases
and more condensed products) and the chemical composition changing (with fewer
aromatics and more aliphatics produced over time). With a fresh activated carbon bed,
a significant proportion of the input HDPE mass was retained within the reactor at the
end of the experimental run. This accumulation of mass, the correlated loss of catalytic
activity of the activated carbon bed, and the decrease in rate of accumulation as the
activity of the bed itself declined, suggested the activated carbon was deactivated
by coking. Indeed, it seems that activated carbon catalyses not only the cracking
of hydrocarbon, but also its dehydrogenation, leaving a carbonaceous remainder that
interferes with the catalytic ability of the bed.
The deactivation process itself was temperature dependent, though in a highly
non-linear fashion, occurring very rapidly at temperatures of 500°C and above, but
significantly slower at 450°C and below. At 450°C the bed was able to process around
3.5 times its own mass in HDPE before any significant changes in pyrolysis product
composition were noticeable. The pyrolysis products were almost entirely unbranched
aliphatic, and aromatic species, with the balance between the two changing over time:
aromatics were observed in large quantities with virgin activated carbon, and these
declined as the activated carbon deactivated. Using a coke bed the pyrolysis products
were comprised almost entirely of linear aliphatics, up until the highest temperatures
tested where aromatic species started to be produced. Generally, as more HDPE
was processed by the activated carbon bed and it deactivated, the characteristics
5.8 Summary 139
of the pyrolysis products in terms of both molecular mass distribution and chemical
composition tended towards those obtained when using the non-catalytic coke bed.
The differences arising from the batch and continuous addition of sample were
investigated using the coke bed, in order to avoid the catalytic influence of the
activated carbon which changed over time. While at 400°C slightly more cracking
of HDPE occurred in batch mode, generally the average molecular mass across all
pyrolysis products was less when operating in continuous mode, i.e., more cracking
occurred with the continuous addition of sample. This was thought to result from
different residence times arising from the fundamental kinetics underlying the two
different addition modes, a conclusion supported by the development of a simple
illustrative mathematical model using first order kinetics. The abrupt introduction
of a large quantity of HDPE in the batch mode caused large quantities of pyrolysis
gas to be rapidly produced; the resulting increase in volume from the solid to gas
phase change rapidly forced pyrolysis gas from the reactor, decreasing residence time.
While this declined exponentially as the quantity of HDPE remaining in the reactor also
declined, the average residence time across a batch run was still shorter in all but the
lowest temperature cases, resulting in the pyrolysis products undergoing less cracking
compared to the continuous case, where the residence time was longer and quickly
reached a steady state governed by the rate of introduction of sample.
While the condensed pyrolysis products resulting from the microwave-assisted
pyrolysis of HDPE using an activated carbon bed in continuous mode initially all lay
within the desirable molecular mass range suitable for use as liquid transport fuels,
as the activated carbon deactivated heavier aliphatics began to form, with the bed
able to process around 3.5 times its mass in HDPE before these became evident. The
relatively rapid deactivation of the activated carbon, and the resulting inconsistency of
the output products beyond this point, calls into question the economic feasibility of
the overall process, unless an inexpensive method of reactivating the activated carbon
could be found.
The in-depth examination of the pyrolysis products’ yields and chemical composition
revealed that while activated carbon is a useful catalyst in the microwave-assisted
pyrolysis of HDPE, it deactivates relatively rapidly in continued use. Thus far, the
data gathered on the catalyst itself has been indirect — resulting from examination
of the products produced using the catalyst. In order to build up a more complete
understanding of the catalytically assisted cracking, and the observed changes in the
activated carbon’s catalytic ability, it was next necessary to examine the activated
140 Continuous Pyrolysis of HDPE using an Activated Carbon Reactor Bed
carbon itself. As such, the remainder of this work investigates the catalytic mechanisms
of the activated carbon accounting for the differences in the pyrolysis process and
produced products that have been outlined over the last two chapters, with a particular
emphasis on the physical and chemical changes in the activated carbon over time that
underlie its observed decline in catalytic activity.
6
Mechanisms of Pyrolysis, Catalysis,
and Deactivation
6.1 Introduction
Over the previous two chapters it has been established that pyrolysing HDPE over
fresh activated carbon produces products with a considerably different chemical
composition, and which are considerably more cracked, than those produced using a
non-catalytic bed. This was achieved by first establishing that this difference existed at
all (chapter 4), and then by attempting a more detailed characterisation by processing
larger quantities of HDPE in a continuous input mode (chapter 5), during which time it
became clear that the activated carbon deactivated as this process continued, with
the resulting pyrolysis products gradually approaching the non-catalytic coke bed
profile over time as this occurred. Having investigated the activated carbon catalysed
pyrolysis of HDPE in some depth, and detailed the changes in the carbon bed and
pyrolysis products (the “what”), it was then necessary to investigate the causes of these
141
142 Mechanisms of Pyrolysis, Catalysis, and Deactivation
changes and the mechanisms that underlie them (the “why” and “how”).
This chapter presents a set of plausible reaction pathways that are able to account
for the species observed in the pyrolysis products; first, those produced with the non-
catalytic coke bed; this is then followed by an attempt to establish reaction mechanisms
and pathways for the catalytic case, deriving these from first principles from the known
and observed structural characteristics of activated carbon.
A similar process is applied to understanding the mechanism underlying the
deactivation of the catalyst, with a detailed examination of the changes in the activated
carbon bed, and the physical, structural, and chemical properties of the retained
fraction informing a thesis that accounts for the deactivation of the activated carbon
over time.
6.2 Pyrolysis Reaction Mechanisms
Any discussion of the pyrolysis of HDPE would be incomplete without dealing with
the reaction mechanisms that underlie its breakdown into smaller hydrocarbons.
Accordingly, this section applies theory taken from the literature, and data gathered
during the course of this work, to describe both the thermal and catalytically enhanced
breakdown of HDPE into useful shorter chain hydrocarbon products.
6.2.1 Thermal Decomposition of HDPE
Thermal decomposition of HDPE proceeds via a free-radical mechanism, with much
of the pioneering work developing the theory of radical-mediated decomposition of
hydrocarbons undertaken by Rice and co-workers in the 1930s and ’40s [229–235].
The following summary applies the current working model of this theory to the
decomposition of HDPE to account for the species observed in the pyrolysis products
over the previous two chapters.
6.2.1.1 Initiation
Beginning with an arbitrary HDPE segment:
C
H
R
H
C
H
H
C
H
H
C
H
H
C
H
H
C
H
H
C
H
H
R (6.1)
6.2 Pyrolysis Reaction Mechanisms 143
Initiation of the thermal degradation of HDPE occurs by random scission: decompos-
ition of hydrocarbons occurs through the homolytic rupture of covalent C–C bonds,
rather than C–H bonds, owing to the lower bond dissociation energy of the former (at
347 kJ/mol compared with 413 kJ/mol) [14, 229]. This results in the formation of two
shorter hydrocarbon chains with free radical termini [230]:
C
H
R
H
C
H
H
C
H
H
C
H
H
+ C
H
H
C
H
H
C
H
H
C
H
H
R (6.2)
However, it has been shown that in an open reaction system where all pyrolysis gas is
continuously removed as soon as it is produced, random scission alone is not able to
reduce the molecular size of hydrocarbon molecules below approximately C72. Below
this point, the energy required to cleave the C–C polymer backbone is greater than
that required to vaporise the molecule itself [236, 237]. Clearly other mechanisms are
necessary to explain the further cracking of hydrocarbon These are thought to involve
free radical hydrogen transfer reactions.
6.2.1.2 Inter- and Intra-molecular Hydrogen Transfer
The probability that the two radicals produced in equation 6.2 collide again is negligibly
small, therefore the free radicals produced in this way stabilise by reacting with
themselves or surrounding hydrocarbon molecules [229]. This can occur through
inter-molecular hydrogen transfer, where an atom of hydrogen is abstracted from a
neighbouring molecule, forming a stable saturated alkane and a secondary radical in
the middle of the neighbouring molecule:
C
H
R
H
C
H
H
+ C
H
R
H
C
H
H
C
H
H
R C
H
R
H
C
H
H
H + C
H
R
H
C
H
C
H
H
R (6.3)
Alternatively, intra-molecular hydrogen transfer may occur (equation 6.4 on the next
page), where the hydrocarbon backbone “coils back” to form a transient 5- or 6-
membered ring intermediate [238], before free radical hydrogen transfer occurs
between carbons in the terminal and 5th positions [239] (though 1,4- and 1,6-
transfer are also possible [240]). These molecules may then undergo consecutive
isomerisations, each time changing the position of the radical on the hydrocarbon
144 Mechanisms of Pyrolysis, Catalysis, and Deactivation
backbone — the likelihood of this phenomenon increases with the chain length as the
polymer gains more “flex” [235].
R
R
(6.4)
The secondary intra-molecular radicals formed through inter- and intra-molecular
hydrogen transfer are thought to be considerably more stable than the terminal radicals
formed initially by thermal scission, and the rates of both of these forms of free radical
transfer are very fast in comparison with the rate of random scission [209, 241],
accounting for the pyrolysis products having a mass less than the C72 limit imposed
by thermal random scission alone.
6.2.1.3 Beta-scission
The intra-molecular radicals formed through inter- and intra-molecular hydrogen
transfer most commonly stabilise through the process of beta scission [181, 235]
(equation 6.5), forming two smaller hydrocarbons: a terminal alkene and a new
terminal radical. This new terminal radical then initiates a new cycle of hydrogen
transfer reactions, and the cracking process continues in a chain reaction fashion until
the resulting molecules are sufficiently small that they enter the gaseous state and
leave the hot zone of the reactor. When an existing alkene is attacked by a radical, an
alkadiene and second radical results [242].
R C
H
H
C
H
C
H
H
C
H
H
C
H
H
C
H
H
R R C
H
H
C
H
C
H
H
+ C
H
H
C
H
H
C
H
H
R (6.5)
6.2.1.4 Termination
The decomposition chain reaction can terminate with the recombination of two radicals
to form an alkane:
C
H
R
H
C
H
H
+ C
H
H
C
H
H
R C
H
R
H
C
H
H
C
H
H
C
H
H
R (6.6)
6.2 Pyrolysis Reaction Mechanisms 145
An alternative termination pathway is through disproportionation, resulting in an
alkane and alkene [177]:
C
H
R
H
C
H
H
+ C
H
R
H
C
H
H
C
H
R
H
C
H
H
H + C
R
H
C
H
H
(6.7)
6.2.1.5 Non-Catalytic Pyrolysis of HDPE
The reactions detailed thus far are thought to be the primary mechanisms for the pure
thermal decomposition of HDPE. These reactions alone are able to account for the vast
majority of species observed in the pyrolysis products produced using the non-catalytic
coke bed, in particular the alkadiene (eq. 6.5), alkene (eq. 6.5, 6.7), alkane (eq. 6.3, 6.6,
6.7) triple peaks described in §4.5.2. Of the two hydrogen transfer mechanisms, intra-
molecular transfer is thought to dominate, especially at higher temperatures, owing to
the hydrogen-rich nature of the HDPE polymer backbone, and the flexibility of the C-C
bonds.
The activation energy of beta-scission is higher than both the hydrogen transfer
and the termination reactions. The proportion of beta-scission reactions will therefore
increase with reaction temperature [173]; this accounts for the observed distribution of
alkenes and alkanes in the pyrolysis products, particularly the temperature-dependent
increase in alkene prevalence.
6.2.2 Cyclic and Aromatic Formation
6.2.2.1 Diels-Alder Reactions
The formation of aromatic and cyclic compounds has traditionally been ascribed to
Diels-Alder reactions [228, 243], schematically shown in figure 6.1 on the following
page. In this thermally driven cyclo-addition reaction, a conjugated alkadiene and
alkene react to form a substituted cyclohexene intermediate. In conventional Diels-
Alder reactions the aromatic then forms from successive dehydrogenation, or loss of
substituted groups, from the cyclic alkene; this dehydrogenation is likely to occur at
least in part through the hydrogen transfer mechanisms already outlined, given the
large number of radicals that will be present at pyrolysis temperatures in the reactor.
146 Mechanisms of Pyrolysis, Catalysis, and Deactivation
R1
R2
+
R3
R4
R2
R4
R3
R1
radical hydrogen
abstraction
R2
R4
R3
R1
dehydrogenation
+ R1 R3 + R2 R4
Figure 6.1: The formation of aromatic species through the Diels-Alder reaction. In traditional
Diels-Alder the aromatic species forms by consecutive dehydrogenation of the cycloalkene,
though this can also occur through radical-mediated hydrogen abstraction, leaving the functional
groups attached to the benzene ring.
Equation 6.8 presents a specific example of the generic reaction presented in figure
6.1, giving the formation of toluene, the most common aromatic observed in the
pyrolysis products:
+ + 4H• (6.8)
In this reaction, toluene is produced from 1,3-butadiene reacting with propene; the four
hydrogen atoms could be abstracted from the cyclohexene ring by radicals, but this
reaction is unlikely to proceed through bi-molecular recombination as no molecular
hydrogen was observed in the produced non-condensable gases. The concentration
of 1,3-butadiene observed in the pyrolysis gas produced using a coke bed increases
with temperature, correlating with the concentration of aromatics observed in the
condensed products. More generally, this scheme is representative of the class of
reactions that is able to account for the bulk of aromatic species observed: benzene
rings with variable length substituted alkyl groups. The terminal alkene reactants
required for this pathway are produced through the beta-scission mechanism already
outlined. In the case of longer alkene reactants, the resulting longer alkyl groups would
not be immune to further cracking and radical attack for as long as they were present
in the reactor. However, the methyl group in the toluene structure is largely immune
to this form of attack, as stabilisation of any radical that forms here, other than by
hydrogen abstraction back to the methyl group, requires the disruption of the benzene
6.3 Catalytic Activity of Activated Carbon 147
ring — a process with a considerably higher activation energy than maintaining the
methyl group; this accounts for the stability of the toluene molecule and its status as
the most abundant aromatic species observed in the pyrolysis products.
6.2.2.2 Radical-Mediated Aromatic Formation
More complex radical-mediated mechanisms are also likely to result in the formation
of aromatic species, and indeed these pathways may represent the dominant mode
of aromatic formation. However, defining these exact reaction pathways is an area
of ongoing research, with a multitude of possible mechanisms proposed to explain
aromatic ring formation. Many of these pathways are bimolecular, deriving from
the combination of short-chain alkenes in the gaseous state, though mono-molecular
pathways are also posited, where a terminal radical coils back and reacts intra-
molecularly [244–247]. Equation 6.9 presents a simple example of cyclisation and
aromatisation derived from a terminal alkene and radical [14]; the existence of these
initial reactants is easily explainable using the mechanisms already discussed.
R1
+
R2
-H•
R2
R1
dehydro-
genation
R2
R1
(6.9)
Indeed, given the stochastic decomposition of long-chain hydrocarbons, and the
plethora of species that result, it is unlikely that formation of aromatic species is limited
to any one particular scheme. It is worth noting, though, that some of these schemes
offer alternative radical stabilisation and chain reaction termination pathways to those
detailed in §6.2.1.4. Given the complexity, uncertainty, and preponderance of these
secondary pathways, cataloguing them lies outside the scope of this work, though the
interested reader is directed toward a comprehensive review by Richter and Howard
[248], which covers some of these schemes.
6.3 Catalytic Activity of Activated Carbon
From the lighter molecular masses and different chemical make-up of the pyrolysis
products produced using an activated carbon bed, it is clear that the cracking of
HDPE occurs more rapidly, and with altered pathways, compared with the use of a
non-catalytic coke bed. The mechanism of activated carbon catalysis of hydrocarbons
148 Mechanisms of Pyrolysis, Catalysis, and Deactivation
(in isolation, and not as a support material in conjunction with, e.g., metal ions) is
generally not well characterised, and suffers from conflicting evidence in the literature.
While the elucidation of a complete catalytic mechanism for activated carbon is beyond
the scope of this work, the subject does bear discussion. As such, several different
possible mechanisms accounting for the observed catalytic activity of activated carbon,
when used in a microwave-heated environment, are presented below. The mechanisms
discussed here are by no means mutually exclusive, and indeed it is quite possible that
they act together in concert to produce the cracking and product profile observed when
using the activated carbon bed.
6.3.1 Catalytic Cracking Reaction Mechanisms
While carbon is usually thought of as an inert material, the surface of activated carbon
has a number of active sites, deriving from unsaturated valencies at the edges of
graphene layers, heteroatoms, and defects within the graphene structure. The quantity
of all of these kinds of active sites increases with the surface area and porosity of a
sample [52], and these sites are likely to underpin activated carbon’s catalytic ability,
with the activity of carbon catalysts linearly correlating with the concentration of these
surface groups [38].
6.3.1.1 Microporous Structure
The very microporous structure of the activated carbon itself is also likely to play a role
in the catalytic activity observed in the formation of the described pyrolysis products,
with micropores themselves showing an ability to stabilise free radicals, and promote
bimolecular reactions in the pore space (such as the radical mediated formation of
aromatic species described previously) [249–251].
6.3.1.2 High Energy “Defect” Sites Facilitating Hydrogen/Radical Transfer
Greensfelder et al. [252] theorised that the high surface area of the activated car-
bon acts to abstract hydrogen from hydrocarbons (generating radicals), and return
hydrogen to free radicals. In other words, the activated carbon acts as a catalyst
to reduce the activation energy of hydrogen and radical transfer reactions. While
Greensfelder et al. proposed no mechanism for this transfer to occur, a possibility arises
when the structure of activated carbon is considered. Semi-amorphous, “disordered”
carbons such as activated carbon consist of layers that exhibit short-range hexagonal
6.3 Catalytic Activity of Activated Carbon 149
graphene lattice structure, but with the “defective” remaining structure resulting in
folds, discontinuities, and irregular morphology, from which the 3D structure of the
pores derives [52].
The defects in the graphene structure, with the associated disruptions in the regular
array of carbon bonds, result in a range of energetic abnormalities such as unsaturated
and “free” valencies, including unpaired and/or delocalised electrons [36, 39]. These
high energy sites may act to donate or accept electrons (radical transfer) and/or
hydrogen (on to and from free valencies). Certainly, the disrupted lattice as described
would be able to accept and donate these species far more readily than a fully saturated
graphene lattice. Indeed, it is widely accepted that the more disorder present in a
carbonaceous material, the greater the exhibited catalytic activity [119, 120, 253],
i.e., it is the structural disorder in the form of high energy active sites and pores
from which the catalytic activity of carbonaceous materials derives. The boundary
cases of highly ordered graphite, diamond powder, and carbon nanotubes show little
to no catalytic activity (the edge area of graphite where unsaturated carbon atoms
occur is small compared with the basal plane, which in any case does not exhibit the
defects discussed), while amorphous “glassy” carbon shows catalytic activity (though
its smaller surface area compared with activated carbon makes it overall a less effective
catalyst per unit mass) [39].
By acting as a hydrogen and/or radical acceptor/donor/transfer agent, the activated
carbon lowers the activation energy for the reactions that are mediated by these
mechanisms, speeding them up considerably. In this method, then, dissociative
adsorption of hydrocarbons on high energy active sites occurs with an initial adsorption
of a molecule onto the surface of the active site, arising from intense van der Waals
forces owing to the near proximity of multiple carbon atoms. Cracking initiation, or
transfer reactions (by which chain reaction propagation occurs), are then facilitated by
the proximity of the hydrocarbon to the high energy sites, which reduce the activation
energy of electron/hydrogen transfer. Given the van der Waals forces holding the
hydrocarbon to the activated carbon surface, it is likely to maintain physical contact
for sufficient time to effect intra-molecular transfer (with a conformation change or
small movement in the hydrocarbon molecule); alternatively, the hydrocarbon may
desorb, with another molecule taking its place, facilitating a potential future intra-
molecular transfer. A final possibility is the molecule progressing from physisorption
of the hydrocarbon to chemisorption, with the molecule bonding to the basal structure
of the catalyst; this would also present a consistent mechanism for deactivation and
150 Mechanisms of Pyrolysis, Catalysis, and Deactivation
elimination of these high energy sites, the possibility of which is explored further in
§6.9.
It is acknowledged that the detailed mechanism for this catalysis is complex and
requires significant additional research before it is completely characterised and
understood. However, this theory is consistent with the current knowledge of the
structure of activated carbon, and with the products of activated carbon mediated
catalytic cracking of HDPE observed in this work.
6.3.1.3 Accounting for Activated Carbon Product Spectrum by Catalytic
Mechanism
The major differences between the products produced using the activated carbon bed
and the coke bed are the increase in total cracking, and the increase in the proportion
of aromatic species. From the mechanisms and research outlined above, it is clear
that activated carbon catalyses dehydrogenation [99], which is a central requirement
in the formation of aromatics through the mechanisms discussed. Generally, aromatic
species are highly stable products, but which have a relatively large activation energy
for the formation from aliphatics. The catalysis of hydrogen and radical transfer would
lower this activation energy considerably and allow the formation of aromatics as a
stable end product, and the consequent higher production of alkanes owing to the
“freed” hydrogen. More generally, an across-the-board enhancement of the radical
and hydrogen transfer mechanisms discussed in §6.3.1 would account for the overall
greater cracking observed when the catalytic activated carbon bed was used.
6.3.1.4 Oxygen Active Sites and Acid-Activated Catalysis
The presence of heteroatoms within the carbon lattice (e.g., oxygen, hydrogen, and
nitrogen) also introduces active sites into the activated carbon structure. Of these,
oxygen active sites are thought to play by far the largest role in catalysis, providing
acid sites that are able to crack molecules through acid-activated catalysis [36]. This
process (using different catalysts such as zeolites) underpins fluid catalytic cracking
(FCC), one of the most important processes in the petrochemical industry, which is used
to convert high molecular weight hydrocarbons into more valuable lighter products.
Acid catalysis involves the transfer and acceptance of protons, as originally outlined
by Whitmore [254] (with contemporary mechanistic theory summarised by Corma and
Orchilles [255], and Cumming and Wojciechowski [256]), as opposed to the radical-
based mechanisms proposed thus far.
6.3 Catalytic Activity of Activated Carbon 151
O
O
O
O
O OH
OHO
Quinone Phenol Carboxyl Lactone Cyclic Peroxide
Figure 6.2: Some oxygen-based functional groups present on the surface of activated carbon.
Adapted from [36].
Oxygen-based surface active sites in activated carbon may arise from the original
oxygen-containing source material from which the activated carbon is derived, through
the presence of oxidising species during activating treatment or post-treatment or, most
commonly, oxygen may be chemisorbed onto high-energy sites on the surface of the
activated carbon after exposure to oxygen-containing species in the atmosphere (e.g.,
O2 or CO2) [52]. At least some of these oxygen activate sites would still be present in the
activated carbon even after the pre-treatment in inert nitrogen atmosphere at 700°C
(these sites requiring temperatures greater than 1000°C before they are completely
removed [52]).
Zawadzki [257] and Ishizaki and Marti [258] used infrared spectroscopy to identify
oxygen sites on the surface of activated carbon; these included carboxylates, phenols,
quinones, and lactones, though many different possible configurations for these oxygen
active sites exist, some of which are presented in figure 6.2.
Szymanski and Rychlicki [114] presented evidence for the centrality of these oxygen-
based surface active sites to activated carbon’s catalytic ability by freshly preparing
samples of activated carbon and treating them with hydrogen at 1223 K to remove
any oxygen sites, then using this as a bed with which to dehydrogenate butan-2-ol.
The activated carbon prepared this way showed little ability to catalyse this reaction.
In contrast, activated carbon prepared in the same way but with an extra treatment
of highly oxidising concentrated nitric acid effectively catalysed both dehydration and
dehydrogenation.
Pereira et al. [38] identified carbonyl/quinone surface functional groups as respons-
ible for the catalytic activity of activated carbon in the oxidative dehydrogenation of
ethylbenzene by first introducing oxygen surface groups by treatment of activated car-
bon with oxygen gas, then selectively removing functional groups through progressive
heating of the carbon to high temperatures. After heating to 1023 K, a temperature
at which only carbonyl/quinone groups remained, the reaction remained catalysed. A
further heat treatment to 1373 K removed all oxygen functional groups, after which the
152 Mechanisms of Pyrolysis, Catalysis, and Deactivation
reaction showed no difference compared with activated carbon that had not undergone
the initial oxygen treatment.
It should be noted, however, that both of these studies involve either the decompos-
ition of an oxygen-containing species, or the undertaking of a reaction in an oxygen-
supplied environment, which may reflect an oxidative mechanism different from the
pyrolysis considered in this work. There may also be differences in the functional
groups of the activated carbon used in these experiments: functional groups formed
on activated carbon in the presence of oxygen at high temperatures are significantly
different from “conventional” activated carbon where any oxygen chemisorption occurs
at room temperature [259–261]. Furthermore, in these studies the oxygenated active
sites were artificially induced, either by treatment with oxygen gas or with nitric acid.
Certainly, the acid sites created with nitric acid are significantly stronger than those
evolving through reaction with oxygen present in the atmosphere [52], as occurs with
conventional activated carbon. As such, the oxygen acid sites present in these samples
may have resulted in additional catalytic ability over and above that naturally occurring
in untreated activated carbon; this additional catalytic ability was then removed with
heat treatment returning the carbon to the base state.
A contrasting perspective is presented by Muradov et al. [39], who removed
oxygen-based surface functional groups from activated carbon by treatment with pure
hydrogen at 1123 K, and noted only a small decline (of around 2% vol.) in the
production of hydrogen from methane over this bed compared with the untreated
equivalent. A far more significant drop in catalyst performance and reduction in yield
(of around 40%) was observed over time, which correlated with the reduction in surface
area of the carbon. Muradov et al. concluded that the catalytic activity of activated
carbon could not be solely attributed to oxygen surface functional groups, and that the
majority of catalytic activity derives from surface defects, dislocations, vacancies, and
other high-energy sites as described previously.
Further evidence towards this conclusion was assembled by Petkovic et al. [123],
who, while decomposing hydrogen iodide to produce hydrogen using an activated
carbon catalyst, found no correlation between numbers of any particular type of
acid site (i.e., carboxylic, lactonic, phenolic, etc) and catalyst activity across several
different types of activated carbon tested. Indeed, the carbon with the highest
measurable total acid sites displayed only intermediate catalytic activity in this context.
Acid catalytic cracking proceeds through proton transfer and the generation of
intermediate carbenium ions, which typically result in high levels of isomerisation
6.4 Microwave-Mediated Catalysis 153
and branched products [256]. The fact that these species were not observed in large
quantities in the products produced from cracking HDPE with activated carbon in this
work, suggests that the acid-catalysis pathway plays a smaller role in comparison to
the radical mechanisms already outlined [252]. This is not to say that acid catalytic
cracking does not occur at all, nor indeed, that surface oxygen active sites cannot
act in non-acidic cracking as previously proposed: they are high energy sites in and of
themselves; however, the evidence presented here indicates that it is unlikely that these
high energy locales consist solely of oxygen-based active sites, and further suggests
that activated carbon catalyses hydrocarbon cracking through a different, or at least
an additional, mechanism to that of traditional commercial petrochemical acid-catalysts
such as zeolite.
6.4 Microwave-Mediated Catalysis
6.4.1 Coupling to Free Radicals
Given their prominence in the pyrolysis process, free radicals will be present in
abundance in the reactor as the breakdown of the HDPE occurs. As charged species,
the microwave field will interact not only with the microwave intermediate bed but
also with the radicals themselves as they are produced. The extra targeted energy
imparted to these molecules by the microwave field may enhance the pyrolysis process,
and contribute to further HDPE breakdown above that of a conventionally heated
configuration; though in order to fully test this a direct comparison of the same reactor
configuration with microwave and non-microwave heating sources would be required,
as proposed in the further study possibilities presented in §7.2.1.
6.4.2 Generation of Microplasmas through Microwave
Irradiation
While heating the activated carbon with microwaves, it was noticed that a plethora
of small sparks were present within the bed (see figure 6.3 on the following page).
These sparks, which have also been observed by other researchers [142–144] were
of extremely short duration but were continually formed throughout all visible areas
of the bed whenever it was irradiated by the microwaves. It is thought that these
electrical discharges, or “microplasmas” were created by the interaction of the rapidly
oscillating electromagnetic microwave field with the activated carbon bed, which
154 Mechanisms of Pyrolysis, Catalysis, and Deactivation
Figure 6.3: Photograph of the activated carbon bed, ready to process HDPE, taken from the
continuous rotary feeder input port. The red colour of the bed is from thermal radiation; the
bright spots are microplasmas of ionised gas induced by the irradiation of the bed by microwave
energy.
contains relatively free delocalized pi-electrons, which experience electromotive force
from the microwave field; this created charge imbalances that were restricted by the
physical boundaries of the carbon particles. In order to equalise the induced charge
imbalance, a small arc (both in distance and duration) would form between the points
of high and low voltage, ionising the gas between the points as charge is transferred.
While these microplasmas were observed whenever the activated carbon bed was
used, they were not seen with the coke bed. This is thought to arise from structural
differences between the two types of particle. The activated carbon had extensive
networks of pores throughout the particle structure — these presented physical gaps
which prevented the flow of charge and allowed voltage to build up to the point at which
it would equalise via a microplasma arc. The size of the activated carbon particles and
their irregular shape would also allow this phenomenon to occur in the gaps between
individual particles. In contrast, the coke particles were not only approximately
spherical with no porous structure, but also considerably smaller (90–250 µm vs.
0.52–1.68 mm), with the size and regularity resulting in sufficient physical contact
6.5 Changes in Porous Structure of Activated Carbon 155
between granules that flow of charge could occur unimpeded, thus never building up
to the levels required to form the microplasma arcs.
The extremely high energy of these microplasmas is evident by their ability to ionise
nitrogen gas (with which the bed was pervaded during the capture of figure 6.3):
the nitrogen-nitrogen triple bond has dissociation energy of 945 kJ/mol. Accordingly,
the temperature of these microplasmas is considerably in excess of that of the bulk
reactor bed, and hydrocarbon localised in this area will be correspondingly cracked to
a greater extent. This effect results from the juncture of the microwave heating mode
and the activated carbon structure, and is not technically a catalytic one, though is
still expected to increase hydrocarbon cracking over and above the coke bed case, and
may enhance the initiation of chain-reaction cracking through the creation of radicals
in the hydrocarbon plasma. The lack of any nitrogen containing species in the pyrolysis
products can be attributed to rapid displacement of the nitrogen by the generated
hydrocarbon pyrolysis gas.
6.5 Changes in Porous Structure of Activated Carbon
Physical adsorption studies were carried out on the activated carbon, as a method
for measuring its surface area and characterising the pore volume and area. This
analysis was carried out in order to understand the physical and structural changes
to the activated carbon occurring as it was used in microwave-assisted pyrolysis, and
to determine whether a relationship existed between these changes and the loss of
catalytic activity of the carbon.
6.5.1 Physical Adsorption Analysis Method
Automatic sorption analysis of the carbon samples was carried out using a Micrometrics
ASAP 2010 Automatic Sorption Analyser (Micrometrics U.K. Ltd, Bedfordshire). The
carbon particles were first degassed, then exposed to varying pressures of nitrogen
gas at a temperature of 77K, with the volume of gas that was adsorbed onto the
surface of the sample at each pressure recorded. Nitrogen adsorption and desorption
isotherms were constructed from the data, which allowed the characterisation of the
porous structure, and calculation of the surface area of the sample according to BET
theory [262]. Pore volume and area distributions were determined according to BJH
analysis [263]. The analytical and technical aspects of this analysis were performed
under the guidance and supervision of Zlatko Saracevic, an instrumentation support
156 Mechanisms of Pyrolysis, Catalysis, and Deactivation
HDPE/Bed = 0.0 
HDPE/Bed = 1.9 
HDPE/Bed = 4.1 
HDPE/Bed = 6.9 
Coke 
0
50
100
150
200
250
300
350
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
N
2
 A
d
s
o
rb
e
d
 (
c
m
³/
g
 S
T
P
) 
Relative Pressure (P/Po) 
Adsoprtion
Desorption
Figure 6.4: Nitrogen adsorption/desorption isotherms from activated carbon and coke reactor
beds. For the activated carbon beds the mass of HDPE processed relative to the bed mass is
indicated by HDPE/Bed ; the pyrolysis of the HDPE occurred at 450°C.
technician in the Department of Chemical Engineering and Biotechnology, University
of Cambridge.
6.5.2 Adsorption/Desorption Isotherms
Figure 6.4 shows examples of the adsorption/desorption isotherms from the nitrogen
sorption analysis of the activated carbon. The isotherms can be classified as type I
according to the standard IUPAC system [264, 265]: the rapid adsorption at low
pressure followed by the flattening out indicates the activated carbon has microporous
properties, with the bulk of the absorptive extent occurring with these small diameter
pores. A small amount of type H4 hysteresis [266] between the adsorption and
desorption isotherms results from capillary condensation of the nitrogen and indicates
the presence of mesopores.
It is clear that the absorptive ability of the activated carbon declines dramatically as
it processes more HDPE, with an approximate 10-fold decrease in nitrogen absorbed
after the activated carbon has processed 6.9 times its own mass. This indicates the
surface area of the activated carbon (including catalytic active sites) accessible to
6.5 Changes in Porous Structure of Activated Carbon 157
0
100
200
300
400
500
600
700
800
0 1 2 3 4 5 6 7 8
B
E
T
 S
u
rf
a
c
e
 A
re
a
 (
m
2
 g
-1
) 
HDPE Mass Pyrolysed (Normalised to Bed Mass) 
400°C
450°C
500°C
600°C
Figure 6.5: Change in surface area of the activated carbon as it processes progressively more
HDPE, at different reaction temperatures. Lines of best fit are fitted to a first-order kinetic model
of deactivation described in the text.
molecules decreases as the bed processes HDPE, as demonstrated with the following
BET surface area analysis.
6.5.3 BET Surface Area
Virgin activated carbon has a BET surface area of around 750–800 m2/g; by comparison,
the non-catalytic coke used in the experiments conducted in this work has a surface
area of just 1–3 m2/g. Figure 6.5 shows the change in surface area of the activated
carbon bed over time as it has processed progressively more HDPE. As expected from
the observed changes in produced pyrolysis products, the process is highly temperature
dependent, occurring extremely rapidly at 600°C, and progressively more slowly as the
reactor temperature decreases. The decline in catalytic activity of the bed correlates
well with the reduction in surface area of the activated carbon, in agreement with
the results of Muradov et al. [39] when cracking methane to hydrogen. Independent
of temperature, the quantity of mass retained in the reactor correlates well with the
reduction in surface area of the activated carbon; given the non-volatile nature of the
158 Mechanisms of Pyrolysis, Catalysis, and Deactivation
deposited mass (see §6.6), this correlation is consistent with the deactivation arising
from coking of the activated carbon pores/active sites.
6.5.3.1 Mathematical Accuracy of BET Model
The BET model of surface adsorption assumes a “monolayer coverage” of molecules,
which does not account for cooperative effects or multilayer absorption that is thought
to occur in the micropores of activated carbons [267]. The calculated values presented
in this work thus represent a monolayer coverage “effective surface area” which,
even though it is acknowledged that multilayer adsorption would be occurring, are
nevertheless regularly used to quantify the surface area of activated carbons [268–
272], and are useful for comparative purposes. Indeed, no adsorption instrument
exists that is able to give a “true” measurement of atomic surface area; rather, it is
adsorption capacity that is always measured and only afterwards converted, through
the selection of an appropriate mathematical function, into a calculated surface area
equivalent [52]. The BET model is simple, effective, widely used, and employed here
with full transparency.
6.5.4 Kinetics of Deactivation/Surface Area Loss
The reduction in surface area of the activated carbon, as a proxy for the deactivation of
the activated carbon, is able to be modelled well using first-order kinetics: the rate of
decrease in surface area, with respect to the (normalised) mass of HDPE processed by
the bed, is proportional to the surface area of the activated carbon at any given time:
dB
dm
= −k.B (6.10)
where B is the BET surface area of the activated carbon, m is the mass of the HDPE
pyrolysed, normalised to the mass of the reactor bed, and k is the reaction rate
constant. Separating variables and integrating gives:
B = B0e
−k.m (6.11)
Curve fitting this equation to the data gives the lines of best fit depicted in figure 6.5.
The temperature dependence of the reaction rate constant was modelled using the
empirical Arrhenius equation:
k = Ae−
Ea
RT (6.12)
6.5 Changes in Porous Structure of Activated Carbon 159
-1.4
-1.2
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1.1 1.2 1.3 1.4 1.5
ln
(k
) 
1/Temperature x 103 (K-1) 
Figure 6.6: Arrhenius plot to determine activation energy of deactivation process. R2 > 0.99
where A is the pre-exponential factor, Ea is the activation energy, R is the universal gas
constant, and T is temperature. Taking the natural logarithm of this equation, gives:
ln(k) = −Ea
R
.
1
T
+ ln(A) (6.13)
Thus, the gradient of the plot of ln(k ) versus T-1 shown in figure 6.6 gives the
activation energy of the deactivation process, at 43.2 kJ/mol. As should be clear
from the derivation thus far, the activation energy derived from this calculation cannot
be compared with the activation energy derived from the Arrhenius equation in its
standard use. Traditional application of the Arrhenius equation relates: the rate of
change in concentration of a species with respect to time; the reaction rate constant —
thought of as the number of collisions between reactants, per second, that result in a
reaction (i.e., a change in concentration of reactants); the pre-exponential factor — the
total number of collisions (successful or not) per second; and e−
Ea
RT — the probability
that any given collision will result in a reaction, with the activation energy determining
the temperature (i.e., thermal energy) sensitivity of this likelihood.
With the derivation carried out in this case the reaction rate constant is the number
of collisions between reactant and activated carbon, per unit mass of HDPE processed
by (an equivalent mass of) the activated carbon bed, that result in the deactivation
of the activated carbon (i.e., the decrease in BET surface area); the pre-exponential
factor is the total number of collisions between reactant and activated carbon per
unit mass HDPE processed by (an equivalent mass of) the activated carbon bed; and
e−
Ea
RT is the probability that any of these given collisions as described will result in a
successful deactivation / loss in surface area, with the activation energy determining
the temperature (i.e., thermal energy) sensitivity of this likelihood. Ultimately, though,
it should be remembered that the Arrhenius equation is inherently an empirical relation
160 Mechanisms of Pyrolysis, Catalysis, and Deactivation
in any case.
The purpose of undertaking this calculation is not necessarily to derive a specific
activation energy parameter for external comparison, but rather to demonstrate the
properties of the underlying deactivation reaction: it is well described with a first-
order kinetic model, and demonstrates temperature sensitivity in a mathematically
consistent manner that is well described by the Arrhenius relation. Indeed, this analysis
suggests that the apparent step change in the rate of deactivation of activated carbon
observed between 450°C and 500°C and described in §5.6 is not in fact derived from
a fundamental change in reaction chemistry, but rather is the result of the increase in
deactivation rate resulting from the exponential nature of the temperature sensitivity
of the reaction.
6.5.4.1 Calculated Deactivation Over Batch Experiments
Using this model of BET surface area decline, it is possible to calculate the theoretical
drop in BET surface area that would have occurred over the course of the batch
experiments throughout chapter 4. Figure 6.7 shows the change in BET surface area
over the course of these experiments, progressively applying the calculated rates of
change in surface area for each temperature. While the deviation in surface area of the
consecutive-use bed does grow over time relative to that of a fresh bed, the overall
magnitude of the deviation is not large, owing to the smaller quantity of material
processed in the batch investigation. As an indication of the magnitude of these
400°C 450°C 
500°C 
600°C 
0
100
200
300
400
500
600
700
800
0 0.05 0.1 0.15 0.2
B
E
T
 S
u
rf
a
c
e
 A
re
a
 (
m
2
 g
-1
) 
HDPE Mass Pyrolysed (Normalised to Bed Mass) 
Consecutive
Single run
Figure 6.7: Calculated change in BET surface area throughout batch experiments. The loss in
surface area of the activated carbon bed throughout the batch experiments, where the same bed
was across all runs, is shown as the consecutive data set; the single run data, where a fresh bed
would have been used for every experiment, is shown for comparative purposes.
6.5 Changes in Porous Structure of Activated Carbon 161
changes, extrapolating backwards from the calculated BET surface area to the mass
of HDPE processed equivalent, for the batch consecutive tests, the change in catalytic
activity of the bed for all of the experiments except the final 600°C case lies within
the first data point collected during the continuous experiments. Within this range of
HDPE processed, the composition of the pyrolysis products produced is similar (see
figure 5.2), i.e., there is little difference in the catalytic activity of the bed. Even for the
600°C consecutive case the calculated BET is very close to the first observed data point
in the continuous case at the equivalent temperature, with a surface area of 652 m2 g-1
vs. 627 m2 g-1.
In fact the change in catalytic activity / BET surface area for the batch consecutive
case is likely to be less than that calculated in this comparison, as this model assumes
the same rate of deactivation across batch and continuous addition modes, where
the reduction in average residence time in the batch case would likely reduce the
deactivation of the bed (owing to less contact time between the HDPE and catalyst
for any given quantity of HDPE processed). Overall, then, the deactivation across the
batch experimental set is unlikely to have been substantial.
6.5.5 Distribution of Pore Volume
Figure 6.8 on the following page shows representative examples of the change in
cumulative pore volume of the activated carbon bed as it processes more HDPE. The
pore size distribution was determined from the adsorption branch of the isotherm,
to avoid artefacts from hysteresis and desorption pore network effects [273, 274].
This analysis confirms the initial impressions garnered from the isotherms, with the
vast majority of the activated carbon’s absorptive ability deriving from microporous
volume (note the steep increase in pore volume that occurs with the smallest diameters
in figure 6.8, a point only emphasised by the fact that the diameter scale here is
logarithmic). The coke exhibits little porosity and no microporous structure. These
results are consistent with other studies characterising the porosity and adsorption of
activated carbon [275–277].
The bulk of the porosity present in virgin activated carbon is 3 nm or less,
approximately twenty times the length of a C-C bond [278]. Interestingly, with
virgin activated carbon, the largest products produced had a carbon number of C20,
suggesting a possible correlation between the size of the pores and that of the final
products. Indeed, microporous carbon has been shown to exhibit molecular sieve
properties — the ability to separate molecules based on size and shape [277, 279, 280],
162 Mechanisms of Pyrolysis, Catalysis, and Deactivation
Coke 
0
20
40
60
80
100
120
140
160
180
110100
C
u
m
u
la
ti
v
e
 P
o
re
 V
o
lu
m
e
 (
c
m
³/
g
) 
Pore Diameter (nm) 
HDPE 
Bed = 0.0 
HDPE 
Bed 
= 1.9 
HDPE 
Bed 
= 4.1 
HDPE 
Bed 
= 6.9 
Figure 6.8: Cumulative pore volume of activated carbon and coke reactor beds, as pore diameter
decreases. For the activated carbon beds the mass of HDPE processed relative to the bed mass is
indicated by HDPE/Bed ; the pyrolysis of the HDPE occurred at 450°C. Pore diameter is presented
on an inverted logarithmic scale.
and this property may also contribute to the final make-up of the pyrolysis products
produced using the activated carbon bed; in particular the fixed range of molecular
masses observed in the products initially produced using activated carbon (i.e., up until
the bed had processed around 3.5 times its mass of HDPE at 450°C).
The loss of porosity and surface area that occurs as the activated carbon processes
more HDPE occurs primarily from the micropores. However, while the bulk of the
loss derives from pores of this size, there is an across the board reduction in pore
volume, independent of pore diameter, roughly in proportion to the volume present. Not
only does this marry extremely well with the first-order kinetic model of surface area
loss developed in §6.5.4, but furthermore it provides an insight into the nature of the
underlying mechanism for the loss: that the loss in pore volume is experienced across
all pore diameters, and in proportion to the pore volume at that diameter, is consistent
with individual stochastic molecular deposition of material across the activated carbon
surface.
6.5 Changes in Porous Structure of Activated Carbon 163
0
100
200
300
400
500
600
700
800
0% 5% 10% 15% 20% 25% 30%
S
u
rf
a
c
e
 A
re
a
 (
m
2
 g
-1
) 
Deposited Mass as % of Reactor Bed 
400°C
450°C
500°C
600°C
Figure 6.9: Temperature dependence of coke deposition, and effect on surface area of activated
carbon.
6.5.5.1 Temperature Dependence of Deposition Distribution
While the preceding description of deposited material accumulating across the entire
surface area of the activated carbon holds true at 400°C and 450°C, at higher
temperatures the relationship between deposited mass and porous surface area does
change somewhat. As seen in figure 6.9, the rate of reduction in porous surface area,
per unit mass deposited, increases with increasing reactor temperature; however, at
high temperatures, the surface area does not initially reduce as much, for the amount
of mass deposited, as at lower temperatures. These observations can be explained by
considering the processes involved in the deposition of coke in porous media.
At low temperatures, where the reaction rate is relatively slow, the pyrolysis gas
perfuses throughout the entire porous structure and deposits evenly across the entire
surface of a particle (as in figure 6.10b). At high temperatures, the deposition reaction
rate is fast compared with the rate at which the hydrocarbon molecules diffuse into
the pores, resulting in deposition occurring more quickly at the mouth of the pores, as
in figure 6.10c. Initially, this results in a higher available surface area under nitrogen
adsorption testing than when coke is deposited across the entire surface area, as the
inner pore network, with little deposition, is still accessible as long as the outer pore
164 Mechanisms of Pyrolysis, Catalysis, and Deactivation
Figure 6.10: Influence of reaction rate and diffusion rate of pyrolysis gas on spatial distribution
of coke deposition. Grey represents deposited material, black represents original activated
carbon particle. Starting with a virgin activated carbon particle (a), if the reaction rate is slow
(i.e., at low temperatures) compared with the diffusion rate, then coke is deposited across the
entire surface of the particle (b). If the reaction rate is fast (i.e., at high temperatures) compared
with the diffusion rate, then deposition occurs more at the mouth of the pores than inside the
particle (c), though over time (t2, d) deposition at the pore mouth will close off the inner surface
completely.
mouths have not been closed off completely. However, the build-up of material at the
pore mouth progresses until access to the inner structure is completely lost (shown
in figure 6.10d) with a corresponding sharp decrease in measured surface area. The
spacio-diffusion aspects of the deposition and deactivation effectively act to magnify
the thermal dependence of the process, though it should be noted that these effects
are incorporated into the empirical rates derived in the deactivation model outlined
previously in §6.5.4.
Evidence supporting the occurrence of this pore-mouth deposition mechanism has
been developed by Gonçalves et al. [281], who measured the adsorption ability of
activated carbon after exposure to propene at various temperatures. Measuring the
ability of the treated activated carbon to adsorb both nitrogen and carbon dioxide
showed that the adsorption of the smaller molecule decreased far less than the larger
one, as the temperature at which the propene exposure occurred increased (and the
rate of reaction dominates over diffusion speed), i.e., a narrowing of the pores occurred
6.6 Chemical Composition of Mass Retained in Reactor 165
at the mouth through which the smaller molecule could still pass and “access”/adsorb
onto the inner porous structure. This effect was particularly noticeable at temperatures
of 500°C and above, supporting the observations of this work.
6.6 Chemical Composition of Mass Retained in
Reactor
6.6.1 Thermogravimetric Analysis
The chemical composition of the deposited material was examined by thermogravi-
metric analysis, with the evolved gases analysed by infrared spectroscopy. This was
intended to determine whether the deposited material was comprised of hydrocarbons
and/or volatile substances that had been adsorbed onto the surface of the bed, or, as
suspected, carbonaceous material resulting from coking and dehydrogenation of the
HDPE pyrolysis products.
6.6.1.1 Method
Analysis of the carbon samples was carried out on a Q500 Thermogravimetric Analyser
(TA Instruments, West Sussex). Samples were heated in the presence of nitrogen
carrier gas from room temperature up to 1000°C at a rate of 20°C/min, and the
change in weight recorded. As there was no obvious method to physically separate
the deposited material from the reactor bed, the two were tested together, with a
bed sample from the end of the series of HDPE pyrolysis experiments conducted
at 450°C analysed (representing the bed with the greatest proportion of deposited
material by mass, at around 25%) compared with a sample of virgin activated carbon
that had not been used to pyrolyse any HDPE. The evolved gases produced during the
thermogravimetric testing were analysed by Fourier transform infrared spectroscopy
(FT-IR) with a Nicolet IS-10 Infrared Spectrometer (ThermoFischer Scientific, USA) to
determine their composition.
6.6.1.2 Thermogravimetric Analysis
Figure 6.11 on the next page shows the thermogravimetric curves of the virgin
activated carbon (0% deposited mass) and the deposited material mix. The decrease
in mass up to 100°C seen in the virgin activated carbon case results from the loss
of water vapour adsorbed by the carbon. This is not seen to any significant level
166 Mechanisms of Pyrolysis, Catalysis, and Deactivation
90
91
92
93
94
95
96
97
98
99
100
0 100 200 300 400 500 600 700 800 900 1000
W
e
ig
h
t 
(%
) 
Temperature (°C) 
0% Deposited Mass
25% Deposited Mass
Figure 6.11: Thermogravimetric curves for virgin activated carbon, and the activated carbon
reactor bed at the end of the series of HDPE pyrolysis experiments conducted at 450°C, by which
time the bed comprised approximately 25% deposited material by mass.
with the deposited mass mix owing to the loss of adsorption capacity and reduction
in surface area, as described in §6.5.3. Between 100°C and around 650°C there is no
significant change in mass for either sample. At temperatures of greater than 650°C
both samples start to decrease in mass at a constant rate. This resulted from the
carbon samples oxidising to CO2, owing to the small amount of oxygen present in the
nitrogen carrier gas. Given a constant concentration of oxygen in the carrier gas, and
this being the rate limiting reactant in the oxidisation process, the absolute rate of
mass reduction should be constant. The difference in rate of reduction between the
two samples can be explained given their differences in initial mass: the absolute rate
of oxidation (as limited by the concentration of oxygen in the carrier gas) is in fact the
same between both samples, with the greater percentage decline in the virgin activated
carbon sample arising from its smaller initial mass.
Tests conducted where the mass decline of the samples was allowed to continue
through to conclusion showed no significant difference between the virgin activated
carbon and the deposited mix, with almost the entirety of the sample mass eliminated
through oxidation, though both left a small quantity of residual ash (<10% wt.).
From the point of view of this analysis, the virgin activated carbon and deposited
6.6 Chemical Composition of Mass Retained in Reactor 167
Figure 6.12: Typical FT-IR spectrum of evolved gases during TGA of carbon samples. Top
figure shows the spectrum over time, with the only significant signal arising at the same time the
carbon begins to decrease in mass as it is oxidised. Bottom signal shows a snapshot of the FT-IR
spectrum at its peak strength; the signal is consistent with that of CO2.
material mix may be considered effectively equivalent, at least to the extent that any
differences between them have been easily explained, and do not arise from underlying
differences in chemical composition. This similarity implies that the deposited material
comprises the same carbonaceous chemical composition, and is not made of volatile
hydrocarbons or any other non-similar material as, even were this matter “trapped”
behind a pore blockage, a deposit with a different chemical composition would exhibit
different thermogravimetric and oxidation patterns.
6.6.1.3 Infrared Spectrometry of Evolved Gases
The infrared spectrometry analysis of the evolved gases is consistent with the account
of the thermogravimetric analysis advanced thus far: figure 6.12 shows the FT-IR
spectrum over the course of the thermogravimetric analysis. The only significant signal
is temporally coincident with the decrease in carbon mass due to oxidation, and was
identified as CO2.
168 Mechanisms of Pyrolysis, Catalysis, and Deactivation
6.7 Electron Microscopy
The activated carbon samples were analysed using scanning electron microscopy
(SEM) to visually investigate any structural changes that occurred as a result of the
deactivation process. This was then coupled to energy-dispersive X-ray spectroscopy
(EDS) in order to perform an elemental chemical analysis of the carbon.
6.7.1 Method
Imaging was performed using an XL30 Field Emission Gun SEM (FEI, Oregon, USA),
operated at 5 kV; in EDS mode the SEM operated at 20 kV, and was coupled with
an Oxford Instruments Si/Li atmospheric thin window detector; data collection and
analysis was performed using Oxford Instruments’ INCA software. The analytical and
technical aspects of this analysis were performed under the guidance and supervision
of Dr Jeremy Skepper, Technical Director of the Multi-Imaging Centre, Department
of Physiology, Development, and Neuroscience, University of Cambridge, whose
assistance is gratefully acknowledged.
6.7.2 SEM Images
Figure 6.13 shows the SEM micrographs of the activated carbon particles. Of necessity,
a limited number of images are presented in this dissertation; however, while carrying
out the microscopy several areas were scanned and examined for consistency. No
significant deviations in structure were observed throughout the microscopy work, and
these micrographs are fully representative of the observed structure.
Broadly speaking, two types of morphology were seen in the activated carbon
particles: planar, relatively flat surface (figures 6.13 a and b), and porous areas that
extended into the particle (figures 6.13 c and d). It should be noted that at this
micrometer resolution the only pores visible will be macropores; micropores, which
account for the bulk of the activated carbon’s porosity and absorptive ability, have a
diameter of 2 nm or less, and are difficult to image without specialised setup including
high-resolution transmission electron microscopy and significant image processing
[269, 282]. Nevertheless, significant differences were observed between the virgin
activated carbon and that which had been used to process HDPE. The virgin activated
carbon had a relatively smooth exterior, with some loose carbon/dust particles evident
on the surface. By contrast, the texture of the used activated carbon (which in this
6.7 Electron Microscopy 169
a) b)
c) d)
Figure 6.13: SEM micrographs of virgin activated carbon (a,c) and activated carbon that has
processed 6.9 times its mass of HDPE at 450°C, resulting in significant surface deposition
(evident from its textural change). While larger macropores are still evident (though reduced
in diameter), the deposition of material on the surface of the carbon would effectively block
micropores (which result from the structure of the surface itself). Images a and b show mostly
planar particle surface and are 75 mm across, images c and d show macroporous structure and
are 160 mm across.
170 Mechanisms of Pyrolysis, Catalysis, and Deactivation
Figure 6.14: Typical energy-dispersive X-ray spectrum of activated carbon sample. Y-axis has
been shortened so that elements other than carbon are visible (though carbon is by far the
dominant element).
case had processed 6.9 times its mass of HDPE and lost most of its porous adsorptive
ability) is quite different, with a rough, amorphous surface as a result of stochastically
deposited material. Given the size and extent of the accumulated deposits, which
cover the entire imaged surface of the carbon particle, it is clear that any microporous
catalytic surface of the original activated carbon particle would be almost completely
subsumed by the deposited material (evident from the surface texture change), with
the chemical properties of the bed reflecting its new surface coating.
6.7.3 EDS Elemental Analysis
EDS analysis showed no substantive differences between the activated carbon and the
deposited material, with multiple physical points compared on both virgin and used
activated carbon samples. Figure 6.14 shows a typical spectrum derived from this
analysis. In all cases carbon comprises by far the main element in the samples. Small
quantities of oxygen are present, deriving from oxygenated surface functional groups
(carbon surfaces are never free from oxygen except under extreme conditions such
as heating to >950°C under vacuum [52]). Small quantities of silicon, sulfur, and
aluminium were observed in some cases, though these were present in both the virgin
and used activated carbon samples so were thought to derive from the bituminous
origin of the activated carbon.
6.8 Attrition and Thermal Cycling
In order to attribute the deactivation of the activated carbon solely to coke deposition,
physical and thermal deactivation methods were considered and tested as potential
contributors to the deactivation process. Given the bed was continually stirred
6.9 Mechanism of Activated Carbon Deactivation 171
throughout the experiments, deactivation through attrition of the activated carbon
particles remained a possibility, and was tested for by running the stirrer with a fresh
batch of activated carbon, regularly measuring the mass of the bed, and taking samples
to test for any changes in surface area; figure 6.15 presents the results of this attrition
test. A very small quantity of extremely fine dust was generated, which left the reactor;
this occurred during the first 20 hours of continuous stirring and accounted for around
0.3% of the initial mass, an insignificant quantity in the context of the large mass
changes observed through coke deposition. The BET surface area varied by around
±2% over the course of the test, but no clear trend was evident over this period; this
variation was thought to represent the precision of the measuring instrument.
The activated carbon beds underwent multiple intense thermal cycles throughout
their use; however, testing conducted on a fresh bed with no HDPE added showed
no significant change in mass or BET after the first burn-off that all activated carbon
underwent before first use, in spite of undergoing six subsequent cycles where the
carbon was heated to 700°C for one hour. These results effectively eliminate attrition
and thermal decomposition as deactivation mechanisms for the activated carbon.
6.9 Mechanism of Activated Carbon Deactivation
From the observed change in product yields, the decrease in cracking, increase in
molecular masses of the pyrolysis products, and the change in chemical make-up of
the products themselves, it is clear that activated carbon rapidly loses its catalytic
-10%
-8%
-6%
-4%
-2%
0%
2%
4%
6%
8%
10%
0 20 40 60 80 100 120 140 160
D
e
v
ia
ti
o
n
 f
ro
m
 A
v
e
ra
g
e
 
Hours  
Bed Mass
Surface Area
Figure 6.15: Testing for attrition by change in bed mass and surface area over time.
172 Mechanisms of Pyrolysis, Catalysis, and Deactivation
ability as it processes HDPE. The elimination of attrition and thermal decomposition as
mechanisms, and the correlation between mass accumulated in the bed and reduction
in catalytic ability, combined with the inability to separate this material retained in the
reactor from the activated carbon bed, points to the deposition of a substance directly
on to the surface of the activated carbon in a way that either eliminates or blocks off the
active sites from which the catalytic ability derives. Thermogravimetric and energy-
dispersive X-ray spectroscopy established that the deposited substance was carbon
— indistinguishable from the activated carbon bed from the point of view of these
analyses. From a chemistry perspective, though, it is clear that the carbon deposited
during the catalytic decomposition of HDPE does not possess catalytic activity itself,
otherwise the cracking process would become auto-catalytic and stabilise rather than
rapidly decline to the level observed with no catalyst present. There must therefore
be a substantial difference in structure of the deposited carbon compared with that of
the activated carbon; the nature of this difference is suggested by the decline in BET
surface area as the bed mass increases. To account for the differences described thus
far, the deposited carbon is likely to have regular structural order, with a resulting low
surface area, and a lack of catalytic active sites.
In a study of the deposits resulting from the catalytic decomposition of methane over
carbon black (another disordered/semi-amorphous carbon), Muradov et al. [39] used X-
ray diffraction to investigate the structure of the combined carbon black and deposits.
The diffraction patterns revealed columnar ordering consistent with a graphite-like
structure (though the observed columnar spacing was slightly further apart than
standard graphite). This ordered structure was not present in the diffraction patterns
in virgin carbon black, confirming the expectation that the carbon deposited in the
cracking of hydrocarbons over catalytic carbonaceous material is structurally ordered,
and thus lacks the folds and irregularities that increase surface area and result in high
energy active sites.
Overall then, hydrocarbons are adsorbed onto the surface of the activated carbon
by the van der Waals forces arising from the folds/pores and the close proximity of
multiple carbon atoms. Here, they may interact with high energy active sites and
undergo catalytic cracking in the manner described in §6.3.1. Given the nature of
the active sites (with free valencies, unpaired electrons, etc.) it is highly likely that in a
percentage of these interactions, the hydrocarbon will progress from physisorption to
chemisorption, with the creation of a covalent bond between the activated carbon and
the hydrocarbon. These bonds may also result from “external” inducement, with an
6.9 Mechanism of Activated Carbon Deactivation 173
already adsorbed molecule subject to radical attack, or just a radical itself adsorbing;
these could then stabilise through interaction and bonding with the high energy active
sites. The end result of such interactions is the creation of an overall more stable
molecular structure, but through incorporation into and alteration of the active site,
the hydrocarbon reduces or eliminates the active site’s ability to catalyse further such
reactions.
Even without specific active site deactivation, the accumulation of non-catalytic
carbon in the bed will reduce the overall catalytic activity of the bed by reducing
the proportion of catalytic material substrate is exposed to (by the end of the 450°C
experiment series the bed comprised 25% accumulated mass), though this reduction
alone is insufficient to account for the observed reduction in bed activity.
As polymer fragments, some of the hydrocarbons that become bound to the surface
of the activated carbon will have long external chains. These chains would still
be subject to further reactions, either cracking and separating off, or undergoing
radical attack, hydrogen abstraction, and incorporation into the basal activated carbon
layer. Over time, the structure of the newly deposited carbon would likely exhibit a
tendency towards the structure of graphene due to its highly stable conformation, but
in any case the hydrocarbons would be likely to undergo further dehydrogenation and
carbonisation. In this way, reacted hydrocarbon deactivates the activated carbon active
sites, while “filling in” and/or blocking pores (or at least the micropores responsible for
catalytic activity; macropore dimensions are not on the same scale as the hydrocarbon
molecules and thus would not present the same van der Waals forces required for
adsorption).
This molecular deposition process would occur wherever the hydrocarbon pyrolysis
gas is able to physically reach. While small molecule hydrocarbons should be able
to perfuse throughout the entire porous structure of the activated carbon, at high
temperatures the time scale of the deposition process will be short compared with
that of the diffusion process, causing deposition to occur more frequently towards pore
mouths, resulting in the deposition distributions previously described in §6.5.5.1, and
the more rapid deactivation as pore mouths become closed off.
The formation of the pore/active-site blocking coke may proceed via the same
mechanism as that of polyaromatic hydrocarbons (PAHs). These compounds are known
to form in the high temperature pyrolysis of polyethylene [283], and were observed
in the condensed pyrolysis products formed with activated carbon in this work. PAHs
are able to form through the sequential addition of hydrocarbons via both Diels-Alder
174 Mechanisms of Pyrolysis, Catalysis, and Deactivation
and radical-mediated mechanisms [244, 248]. Growth proceeding in this way would
mimic the underlying activated carbon graphene structure but without the structural
defects, proceding in a piece-wise fashion from single rings into large networks that
would blanket the activity of the underlying active sites. Whatever the mechanism, it
apparently requires catalytic initiation at the temperatures examined in this work, as
the addition of mass to the activated carbon bed declines rapidly in correlation with
its decline in catalytic activity. While PAH deposition would be sufficiently regular that
the new material would not add micropores or catalytic high energy sites, it is in any
case clear that the deposited material does not contain these, or at least if it does they
are present in far smaller quantities than in virgin activated carbon, and would become
deactivated through the same mechanism as in the original sites, resulting in an overall
exponential decline in number.
It is also worth noting that substances that promote acid-based catalysis such
as zeolite (and which function with a mechanism similar to the oxygen-based acid
active sites discussed in §6.3.1.4) also experience coking when cracking hydrocarbons.
Overall this coking is not dissimilar to the process described here, though the
underlying mechanisms proposed to account for the coke accumulation (of which there
are several owing to ongoing research in this area) proceed along acid/base hydrogen
transfer pathways [256, 284–286].
The mechanism for deactivation proposed here, encompassing modification of high
energy active sites and addition of ordered carbonaceous matter, accounts for all
observations regarding the deactivation of the activated carbon — an increase in
mass that reduces as the activated carbon deactivates, the carbonaceous elemental
composition of the deposited material, the reduction in surface area/pore volume that
accompanies the deactivation — and acts as a logical extension of the argument already
advanced for the catalytic activity described in §6.3.1.
6.9.1 Carbon Nanofilament Formation via Microwave-Induced
Plasma
While the microwave-induced microplasmas discussed in §6.4.2 enhance hydrocarbon
cracking, they are also likely to enhance the deactivation of the activated carbon. When
Fernendez et al. [144, 287] pyrolysed methane over carbon using microwave heating,
a microscopic analysis of the bed showed the growth of carbon nanofibres. Neither the
microplasmas nor the nanofibres were observed with conventional electric heating. It
6.9 Mechanism of Activated Carbon Deactivation 175
Figure 6.16: Carbon filament formation resulting from the creation of microplasmas in the
hydrocarbon pyrolysis gas. The formation shown in this image is approximately 2 cm across.
is thought that the plasmification of the pyrolysis gas occurring in these microwave-
induced plasma arcs is able to dehydrogenate the hydrocarbon and cause deposition.
While no fibres as such were observed in the SEM images of the activated carbon
used in this work, this is because the bed was continuously stirred, resulting in random
deposition at distinct sites, rather than the continual growth that could occur with a
static bed. Dramatic visual evidence that these nanofibres do occur in the apparatus
used throughout this work is presented in figure 6.16. In this case, an undetected
arc formed in the waveguide entering the reactor, a space in which pyrolysis gas was
able to freely pervade during operation. Over time, the arc induced the deposition of
a complex network of carbon fibres. Once initially seeded, the distance between the
opposite wall and the conductive tubule growth site was less than any other point on
the otherwise parallel waveguide walls, so the arc continued in this location (the lesser
distance requiring the least voltage to breach and ionise). The branching nature of
the formation results from the fluctuations in the plasma stream as it traverses the
constantly varying path of least resistance through the turbulent atmosphere.
Carbon fibres have a very low surface area, and regular structure compared with
activated carbon [39, 288], i.e., they are not a source of high energy active sites.
Microwave-microplasma induced carbonisation and deposition is thought to contribute
to the deactivation of the activated carbon described above, though is certainly not the
sole mechanism, given the deactivation observed by other researchers not operating
with microwave-mode heating [119, 120].
176 Mechanisms of Pyrolysis, Catalysis, and Deactivation
6.10 Summary
The mechanisms outlined in this chapter fundamentally underlie the entire body of
work presented in this dissertation. The non-catalytic pathway of HDPE pyrolysis is
thought to proceed via a radical-mediated mechanism, these being created with an
initial random thermal scission, then going on to react in a chain fashion via intra-
and inter-molecular hydrogen transfer and beta-scission. Aromatics are thought to
form through Diels-Alder and/or radical-mediated mechanisms, both of which proceed
via cyclisation, with aromatisation occurring through subsequent dehydrogenation
pathways. These reactions account for almost the entire gamut of species observed
in the pyrolysis products produced using the coke bed, though of particular note is the
ability of these pathways to account for the alkene, alkadiene, alkane triples that make
up so much of these products.
The catalytic pyrolysis of HDPE is thought to derive from high energy active sites
on the accessible surface of the carbon, that result from defects, folds, and disorder
that alters the structure of the activated carbon away from that of graphene. These
sites include unsaturated and free valencies, including unpaired and/or delocalised
electrons, which may act to accept, donate, or transfer electrons and/or hydrogen
(radicals) between species and the activated carbon bed. The catalytic mechanism
may also proceed through acid-catalysis due to oxygen-based functional groups, though
evidence for this is mixed.
The generation of microplasmas through microwave irradiation of the activated
carbon bed may also contribute to cracking of the HDPE owing to the high temperature
of the plasma. This is not a catalytic effect per se, though, and instead arises from the
interaction between the microwave heating mode and the physical structure of the
activated carbon. The microplasmas were not observed when using the coke bed.
The surface area and pore volume of the activated carbon declines as it processes
HDPE and accumulates mass. The loss in surface area, as a proxy for deactivation of
the carbon, was modelled with first order kinetics with a good resulting fit. This model
was able to explain the temperature dependence of deactivation in terms of exponential
temperature sensitivity as described by the Arrhenius equation. This model was then
used to calculate the amount of deactivation that would have occurred throughout the
course of the batch addition experiments. While some deviation from using a fresh bed
every time was noted, overall the catalytic reduction would not have been large, given
the quantity of material processed.
6.10 Summary 177
Generally the accumulation of mass in the bed correlated well with the reduction in
surface area, though at higher temperatures this occurred to a lesser extent initially,
but then progressed at a more rapid rate. These results were thought to arise from the
increase in reaction rate at high temperatures, such that the hydrocarbon would react
and deposit on the surface of the activated carbon before it was able to diffuse very far
into pores. As a result, deposition occurred more often at the mouth of pores; initially
small molecules could still get through the narrow mouth and access the pore volume
inside, accounting for the higher initial measured surface area; however, this was soon
followed by the closure of pores and rapid fall-off in measured surface area.
The chemical composition of the deposited material was confirmed to be coke
through thermogravimetric analysis coupled with FT-IR, and subsequent elemental
analysis using SEM-EDS. The deposited material was indistinguishable from the
activated carbon using these techniques. In spite of establishing the deposited mass
as carbon, there was still an obvious decline in catalytic activity as it was deposited,
indicating that the chemistry/structure of the deposited material was significantly
different to that of activated carbon. The observed reduction in surface area suggested
that the deposited material exhibited regular structure and did not contain the defects,
folds, and irregularities that act as high energy catalytic sites for activated carbon.
This theory was in agreement with Muradov et al. [39], who used X-ray diffraction to
show graphite-like structure in carbon deposited from catalytic cracking of methane
over carbon black.
From these results it was concluded that deposition/deactivation likely proceeded
initially from physisorption of hydrocarbon pyrolysis gas through van der Waals forces
from pore walls, then proceeded through reaction with the high energy active sites,
resulting in the creation of covalent bonding and the integration of hydrocarbon into
the base activated carbon structure (where it underwent subsequent dehydrogenation
and carbonisation). By incorporation into, and alteration of the active site, this mode
of interaction depletes the number of active sites, fills in/blocks pores, and reduces the
ability of activated carbon to catalyse further such interactions.
The mechanisms described in this chapter account for the products observed in
the pyrolysis of HDPE, both with standard microwave-assisted pyrolysis and when
using a catalytic activated carbon bed. By combining theory from the literature with
observed changes in chemical and structural properties of the activated carbon as it
was used to process HDPE, it was possible to develop a comprehensive mechanistic
theory that accounts for both the catalytic properties of the activated carbon, and its
178 Mechanisms of Pyrolysis, Catalysis, and Deactivation
deactivation over time. Having advanced mechanisms to explain the observed data
collected throughout this work, this chapter concludes the analysis portion of this
dissertation. Nevertheless, the material presented in this body of work raises a number
of avenues of potential further study which, along with a broad concluding overview,
are presented in the following and final chapter.
7
Project Overview and Further Study
7.1 Project Overview
This dissertation began by introducing the fundamental principles of the microwave-
assisted pyrolysis of plastics, and presenting the concept of combining the microwave-
receptive properties, and relatively unstudied catalytic abilities, of activated carbon for
use as a reactor bed in this process, with the intention of producing pyrolysis products
with a narrow and useful range of molecular masses. A thorough review of activated
carbon, microwave heating, and the wider pyrolysis field established that this avenue
of research was both novel, and of sufficient potential to warrant further study.
7.1.1 Development of Microwave-Assisted Pyrolysis Equipment
Considerable effort was put into developing safe and reliable apparatus in which to
carry out the microwave-assisted pyrolysis of HDPE. An existing reactor that had been
used to conduct pyrolysis on liquid hydrocarbons was refurbished for this purpose.
A number of improvements to the reactor were made, including the simplification of
179
180 Project Overview and Further Study
the power transmission system — reducing reflected power and eliminating points
of failure; the addition of a three-stub capacitive tuner to the transmission pathway
to match the impedance of the “load” (the reactor bed) with the magnetron source
— thus maximising power transmission; and the design and fabrication of a silicone
window and support structure that allowed microwaves to pass through to the reactor,
but did not allow damaging pyrolysis gas or heat to return back to the magnetron.
A gas-tight method for adding HDPE pellets to the carbon bed when it reached the
target temperature was required, and a plunger-like sample injection system was
developed for this purpose. The development of suitable apparatus with which to
condense produced pyrolysis gases resulted in several iterations based around Liebig
condensers that ultimately proved inadequate owing to the pyrolysis products being
carried through the condensing apparatus by the constant flow of gas and not settling
on the condenser surface. The development of a dry ice condenser through which the
pyrolysis gas percolated provided an effective solution to this issue. The entire system
was controlled by a computer system based around a PID feedback loop; this was able
to vary the power of the magnetron to maintain the reactor at the desired temperature,
as well as performing safety monitoring and stopping power to the magnetron in
the event that any arcing was detected, a process it performed far more rapidly and
effectively than any human controller.
7.1.2 Exploratory Experiments with Batch Addition of HDPE
A series of exploratory experiments were conducted pyrolysing HDPE that was added
in a single batch to an activated carbon bed over a range of temperatures from
400–600°C. Equivalent control experiments were also conducted using a non-catalytic
coke bed. The pyrolysis products produced using the activated carbon bed underwent
significantly more cracking than those produced with the non-catalytic coke bed, with
the condensed products approximately three times lighter on average (with an average
carbon number of 10.0–11.5 vs. 25.0–33.0), indicating that activated carbon could
be used as an effective catalyst in this application. The products produced using
the coke bed were almost entirely linear aliphatics, while the condensed products
produced using the activated carbon bed showed a considerable component of aromatic
species (36–45%). The hydrogen that was “freed” through the formation of aromatic
species, with lower H/C ratios, resulted in a higher concentration of alkanes in the
non-condensable gases.
The range of molecular masses produced when using the activated carbon bed was
7.1 Project Overview 181
much smaller than when the coke bed was used; indeed, the condensed products
resulting from the activated carbon bed shared the same mass range as that of the
liquid transport fuels petrol and diesel. The aromatic portion of these condensed
products consisted of benzene and substituted derivatives (e.g., toluene, ethyl-benzene,
and xylene), substantially valuable products in their own right. The non-condensable
gases resulting from both beds were relatively similar (aside from the greater alkane
proportion in the activated carbon derived gases already mentioned), the species in
this fraction being limited by the range of hydrocarbons in the gaseous state at room
temperature. These short-chain aliphatics could be used as fuels (e.g., natural gas,
LPG) or feedstock for a range of chemical processes including fresh HDPE synthesis
(the gases included a high concentration of ethene, the original HDPE monomer).
Overall, the entire range of products produced with the activated carbon bed had a high
potential economic value, while a large portion of the condensed products produced
using the coke bed were longer-chain waxes that would require further cracking before
they could be used.
7.1.3 Continuous Throughput Pyrolysis of HDPE
Following from the positive results from the batch input study, a second study was
conducted to assess the pyrolysis process in greater detail, processing larger quantities
of plastic with a continuous throughput of material, a mode any potential commercial
plant would be required to operate in for capacity and efficiency reasons. A continuous
rotary feeder was designed and built for this purpose which provided a constant flow
of plastic pellets into the reactor during operation, while maintaining containment of
the pyrolysis gases and microwaves.
Because of the larger quantities of HDPE pyrolysed with the continuous rotary
feeder, and as a result of conducting a number of experiments in sequence, it soon
became clear that the nature of the pyrolysis products produced with the activated
carbon bed was changing over time, with the cracking ability of the bed reducing, and
a significant proportion of mass accumulating in the activated carbon bed with each
experiment. Both the average molecular mass, and the range of molecular masses
of the pyrolysis products increased over time, with a reduction in the proportion of
non-condensable gases and an increase in condensed products (themselves heavier).
The rate of this catalytic deactivation was highly temperature dependent, occurring
very quickly at temperatures of 500°C and over, and much more slowly at and below
450°C. The condensed products produced using the activated carbon bed with the
182 Project Overview and Further Study
continuous addition of HDPE initially had an even higher aromatic content (61–85%)
than those produced with the batch addition, though this declined over time as the bed
processed more HDPE, with a corresponding increase in aliphatic content. Overall, as
the activated carbon deactivated, the resulting pyrolysis products came to increasingly
resemble those produced using the coke bed.
Aside from the catalytic activity of the activated carbon, the primary influence on
the make-up of the pyrolysis products was the temperature at which the pyrolysis
was carried out. Increased temperature had the effect of not only increasing the
rate of cracking, but also reducing the residence time of the input mass as the
pyrolysis products entered the gaseous state and expanded in volume considerably,
generating pressure which drove them out of the reactor. The balance of these
two factors determined the total amount of cracking that the products underwent:
from 450–600°C the overall molecular mass of the combined condensible and non-
condensable gases decreased with increasing temperature, implying the increase in
cracking rate dominated. Decreasing the temperature below 450°C, the overall average
molecular mass decreased again as the time products spent in the hot cracking zone of
the reactor increased.
While the economic value of the products produced in the continuous mode was
much the same as those produced in the batch mode, as a result of the catalyst
deactivation these products changed over time, presenting a potential barrier to
producing a saleable product due to the variability of the output. However, at
lower temperatures there was an initial period during which the condensed products
remained entirely within the liquid transport range: at 450°C this remained the case
until the activated carbon had processed approximately three times its mass of HDPE,
after which time it came to include less valuable heavier waxes.
The differing modes of HDPE input also affected the pyrolysis products: those
produced with the batch input of HDPE underwent less cracking and encompassed a
wider range of molecular masses. This was thought to arise from the different rates of
sample introduction, the sudden bulk introduction of HDPE in the batch mode resulting
in an “explosive” rush of pyrolysis gases being initially produced, and a corresponding
large pressure that forced the gases out of the reactor more rapidly; this then declined
over time as the mass of material undergoing pyrolysis in the reactor reduced. With
the continuous introduction of HDPE, the production of pyrolysis gas quickly settled
to a constant rate that was lower than the initial explosive exit rush from the batch
case. The shorter and more variable residence time in the batch mode explains the
7.1 Project Overview 183
heavier and more varied products, as well as the differences in chemical composition,
with the greater proportion of aromatics in the continuous case arising from the longer
time the pyrolysis products spent in the hot zone of the reactor (aromatic formation
being time-dependent). The effect of addition mode on reaction rate and residence time
was demonstrated with an illustrative mathematical model incorporating the kinetics
behind the two input modes.
7.1.4 Mechanisms of Pyrolysis, Catalysis, and Deactivation
By combining theory from the literature with the chemical profile of the produced
pyrolysis products, reaction mechanisms accounting for the observed species were
advanced. Non-catalytic pyrolysis was thought to proceed via a radical-mediated
pathway, with initiation occurring by random thermal scission of the polymer backbone,
and the resulting highly unstable terminal radicals partially stabilising through intra-
and inter-molecular hydrogen transfer. Both of these pathways result in the formation
of a more stable intra-molecular radical, while the latter of these also produces an
alkane. These intra-molecular radicals stabilise through the process of beta-scission,
resulting in the formation of an alkene, and a new terminal radical that initiates a new
cycle of hydrogen transfer in a chain reaction. Termination of this reaction cascade
occurs through recombination of radicals, or disproportionation (resulting in an alkane
and an alkene). The formation of aromatics was thought to occur through Diels-Alder
type reactions (where a conjugated alkadiene and alkene react to form a substituted
cyclohexene intermediate, which is then dehydrogenated), or more complex radical
mediated pathways, where cyclisation and dehydrogenation also occur.
The catalytic activity of the activated carbon was thought to originate from high-
energy active sites on the surface of the carbon that are created by defects that skew
the atomic carbon structure away from a regular graphene lattice. The resulting
folds and pores are responsible for the very high surface area of the carbon, but
also create unsaturated and free valencies, including unpaired and/or delocalised
electrons, which act to accept, donate, or transfer electrons and/or hydrogen (radicals)
between hydrocarbon species and the activated carbon bed. By facilitating these
transfers the activated carbon lowers the activation energy for radical-mediated
cracking as described above, and other reactions such as the formation of stable
aromatic molecules. Oxygen-based active sites may also play a role in the catalytic
properties of the carbon through acid/base catalysis.
The physical properties of the activated carbon as it underwent deactivation were
184 Project Overview and Further Study
investigated through nitrogen sorption analysis, which revealed a dramatic decline in
surface area as mass accumulated in the carbon bed and it deactivated. This decline
could be modelled well using first order kinetics, with the temperature dependence
fitting the Arrhenius equation, and accounting for the large increase in the rate of
deactivation observed at higher temperatures. The bulk of the porosity present in
the activated carbon was contributed by pores of <3 nm, approximately twenty times
the length of a C-C bond [278], suggesting that an element of shape/size selectivity
may have contributed to the narrow range of the hydrocarbons produced with virgin
activated carbon.
The composition of the mass deposited in the activated carbon bed was confirmed
to be carbon itself through a combination of TGA (with FT-IR of the evolved gases) and
SEM-EDS analyses. Deactivation was thought to occur through hydrocarbons forming
covalent bonds with the activated carbon bed in a percentage of the interactions with
the high-energy active sites, incorporating into the structure and thus deactivating
the carbon by “filling in” the pores, and eliminating the sites that mediated the
electron/hydrogen transfer. The bonded hydrocarbons would then be dehydrogenated
by the abundant radicals in the hydrocarbon gas resulting in an inert carbon layer
without the defects from which the activated carbon gained its catalytic ability.
The microwave heating method caused the formation of “microplasmas” in the
activated carbon — charge imbalances created by the electromagnetic field that were
restricted by the physical boundaries of the carbon particles, and which resulted in
a high temperature plasma spark to equalise the induced voltage. It is likely that
cracking occurred at a greater rate in these microplasmas, and the observation of
the creation of carbon nano-fibres by microwave-induced plasma suggested that these
microplasmas also caused rapid dehydrogenation, likely contributing to coke formation
and the deactivation of the activated carbon. These microplasmas were not observed
in the coke bed, owing to its lack of porous structure, and smaller particle size with
corresponding greater electrical continuity.
7.2 Further Study
As a study that was primarily exploratory in nature, this work leaves plenty of scope for
future work to be carried out to further the understanding of the microwave-assisted
pyrolysis process and the use of activated carbon as a catalyst. Indeed, the results
7.2 Further Study 185
presented in this study suggest a number of potential avenues of exploration, which
are presented in the following section.
7.2.1 Isolation of Microwave Effects
While microwave-assisted pyrolysis results in different heating profiles and character-
istics to conventional heating, as well as additional phenomena in the reactor bed in the
form of microwave-induced microplasmas, further research is necessary to definitively
attribute any observed differences in pyrolysis products and mechanisms to the
microwave-heating mode alone, and controversy remains around whether microwave
heating results in non-thermal effects arising from interaction of the electromagnetic
field with reactant molecules.
To assess these effects on the pyrolysis of plastics, research could be conducted
in a reactor that incorporates both electric and microwave heating mechanisms; with
suitable control software it should be possible to ensure that the heating profile for both
of these methods is the same. With all else being equal (reactor size, configuration,
condensing apparatus, etc.), it would be possible to isolate any differences observed in
pyrolysis process and products to the microwave-heating mode itself. Such a reactor
would be able to conclusively determine whether or not microwave heating, and the
associated microplasma phenomenon, has an impact on any process carried out within
it, including the pyrolysis of HDPE, and the deactivation of activated carbon as posited
in §6.9.1. If additional benefits were recognised from such a study it might be possible
to optimise future reactor and bed configurations to enhance these effects.
7.2.2 Other Catalysts
While this study represents the first application of microwave heating to the catalytic
degradation of HDPE, it has been limited to one catalyst type. There is a large amount
of potential to investigate the decomposition of plastics using other catalysts and
microwave heating. It would be particularly interesting to conduct a comparative study
using zeolites and conventional FCC catalysts and investigate the products resulting
with these catalytic beds. The use of microwave heating in these contexts has the
potential to provide efficiency gains, either from non-thermal interactions or simply
from the direct application of energy to microwave-receptive catalysts without heating
intervening space.
186 Project Overview and Further Study
7.2.3 Composition of Deposited Carbon Material
While the deposited material that deactivated the activated carbon was identified as a
low surface area form of carbon, further study could reveal the structure and form that
this deposited material takes, and thus provide additional insight into the deactivation
reaction mechanisms.
X-ray crystallography is an analytical technique that could be potentially employed
for this purpose, with the ability to identify the presence of regular crystalline graphite-
like carbon that is not present in the disordered structure of activated carbon [39]. As
an alternative or additional tool, Raman spectroscopy is able to distinguish between
different structures of elemental carbon with shift bands identifiable for single crystal
graphite at 1580 cm-1, imperfect graphite and disordered carbons at 1620 cm-1, and
other structural carbon imperfections at 1360 cm-1 [289, 290]. This technique is also
able to provide a quantitative analysis of the relative proportion of these different
structural forms.
Using these methods to compare activated carbon deactivated by pyrolysis in
a conventionally heated reactor with that carbon deactivated through microwave-
assisted pyrolysis could elucidate differences in the deposited material arising from
these different heating mechanisms, for example, whether deposition occurred in a
different fashion as a result of the activity of microwave-induced microplasmas.
7.2.4 Regeneration of Activated Carbon
Given the expense of sourcing activated carbon, and the volume of carbon that would
be required to sustain a plastics pyrolysis plant on an industrial scale, the microwave-
assisted pyrolysis of plastics using an activated carbon bed with a single use cycle is not
likely to be economic. However, deactivation through the pyrolysis of HDPE need not
be the final end to the catalyst. Activated carbon can be regenerated, and future work
could examine this possibility in detail. Microwave heating has been used to regenerate
activated carbon when it has been depleted through the adsorption of pharmaceutical
contaminants [291–293], dyes [294–296], and volatile organic compounds [297–299].
However, in these cases the deactivation occurred by adsorption of molecules at low
temperatures, and not by the pyrolysis of hydrocarbons at high temperatures where a
significant mass of material is deposited and covalently bonded to the carbon surface.
Nevertheless, the fact that microwave heating has been successfully used to “activate”
carbon in the first place [268, 300–302] suggests that even the covalently bonded
7.2 Further Study 187
90%
91%
92%
93%
94%
95%
96%
97%
98%
99%
100%
0
50
100
150
200
250
300
350
400
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
M
a
s
s
 o
f 
B
e
d
 
B
E
T
 S
u
rf
a
c
e
 A
re
a
 (
m
2
/g
) 
 
Treatment Time (hours) 
BET Surface Area
Mass of Bed
Figure 7.1: Activated carbon steam reactivation at 700°C using existing microwave reactor.
deactivated carbon might be reactivated in-situ using the same apparatus with which
it was deactivated in the first place.
7.2.4.1 Preliminary Reactivation Data
A preliminary investigation using the existing apparatus and the depleted activated
carbon from the 450°C series of continuous addition experiments was carried out,
heating the carbon bed to 700°C, and adding water to the bed at a rate of 3.6 mL/min.
As seen in figure 7.1, a partial reactivation was achieved, with the activated carbon
BET surface area increasing from 26 m2/g to 368 m2/g after four hours of this treatment
(compared with around 800 m2/g for virgin activated carbon). The reactivated carbon
was able to produce pyrolysis products with the same amount of cracking as the original
activated carbon achieved with this surface area. The reasonably large increase in
surface area with a relatively small change in bed mass (6% loss, compared with the
25% increase that deactivated the carbon in the first place) suggests that this process
may not only be “burning off” deposited material, but actually creating new porous
structure.
7.2.4.2 Future Reactivation Work
While this initial study represents an initial proof-of-concept, the reactivation process
would need to undergo considerable further development before it became a useful
approach to dealing with the deactivation issue. The initial results show incomplete
188 Project Overview and Further Study
reactivation, with the rate of increase in surface area declining over time. Furthermore,
the process is slow, taking considerable time and energy. In this investigation the
reactivation was carried out at 700°C, the maximum safe operating temperature for the
reactor; however, this is below the 800–1000°C that the activation process is usually
carried out at [275]. It should also be noted that this study was carried out with
activated carbon that saw substantial use before any reactivation attempt was made.
It may be that the longer the coke remains at high temperature within the reactor, the
more “enmeshed” it becomes with the catalyst as it, for example, bonds to multiple
atoms in the carbon bed. Further experimentation at higher temperatures, at earlier
stages of deactivation, and with different oxidising gases such as CO2, may be able to
demonstrate the speed and effectivity improvements necessary to make this a viable
process.
If the reactivation could be made efficient and rapid, it could be incorporated into a
side-by-side regenerator, where the carbon is continually regenerated in an integrated
online facility, as occurs in FCC plants. While the comparison is apt, this would not
simply be a reinvention of current FCC, though it would be interesting to examine the
possibility of employing activated carbon in place of current catalysts in an FCC plant.
As previously argued, activated carbon catalysis proceeds via a different mechanism
to current FCC catalysts, and in any case the profile of the products produced using
activated carbon in this work differ considerably to those produced using standard FCC
catalysts, with a much higher aromatic content of up to 85% (compared with 10–26%
typical in FCC plants [223, 303]). Accordingly, a novel process using activated carbon
would likely focus on the production of aromatics as an end point. The accumulation
of carbon in the bed, with its subsequent activation in the regenerator, would allow
for activated carbon to be a secondary output stream of this process. Overall, online
reactivation would add considerable complexity and expense to the construction costs
of any new plant, and the economic implications of such a development would have to
be carefully considered before it was undertaken.
7.2.5 Modelling of the Pyrolysis Process
As has been demonstrated in this study, the products resulting from the pyrolysis of
plastic are the result of a complex interplay between factors such as mode and rate
of mass addition, heat and mass transfer, reactor temperature, reaction rate, and
residence time. The parameter space that results from the permutations of these
factors is extremely large. However, it should be possible to develop a model of the
7.2 Further Study 189
pyrolysis process that incorporates the known kinetics and mechanisms of pyrolysis,
along with physical heat and mass transfer parameters specific to a reactor, and which
would be able to predict the resulting product in a given configuration. A robust model
with such a predictive capability would be of enormous use in terms of optimising the
production of a desired product, and in the scale-up and design of future pyrolysis
systems.
7.2.6 Future Reactor/Condensing Systems
While the condensing system used in this study performed reasonably well for the
laboratory-scale work conducted here, it presented issues with blockages when high
molecular mass products were produced, and in any case would not scale beyond small
scale experiments owing to the expense and infeasibility of using dry ice on a larger
scale. Two alternatives to the dry ice percolating condenser are presented here, which
have the potential to be used in any future microwave-assisted hydrocarbon pyrolysis
work.
7.2.6.1 Fractionating Column
Given that one of the issues arising from the pyrolytic breakdown of plastics is the
wide range of masses in the produced products, a condensing system with the ability to
separate these products into similar fractions by mass would have considerable value.
The gold-standard in industry for the separation of hydrocarbons based on molecular
mass is the fractionating column. The selective targeting ability of microwave heating
should make it possible to combine the fractionating column concept with a microwave
reactor, a concept illustrated in figure 7.2. In this scheme, similar microwave
generating equipment could be used to that employed in this study (1), with a carbon
bed (2) and feeder (3) also operating on the same principle. Pyrolysis would occur in the
bed as usual, but rather than having gases leave the reactor and travel into a separate
condensing system, the reactor itself could be extended upwards into a fractionating
column, with the heat required for operation generated from the microwave-heated
carbon bed. Volatile pyrolysis gases would rise upwards, and condensing liquids flow
downwards, with condensed substances reaching an equilibrium point on a series
of trays (4) spaced vertically throughout the column, from which the products of a
particular molecular mass can be tapped. “Bubble caps” on the trays help to ensure
good contact between up-flowing vapour and down-flowing liquid. Vapour reaching
the top of the column could enter an external condenser (5), where any recovered
190 Project Overview and Further Study
outlet
top vapour fraction
to condenser
reflux
outlet
1
2
34
5
6
Figure 7.2: Schematic of fractionating column reactor. Numbered components are described in
the text. Fractionating column graphic modified from [304] under creative-commons licensing.
liquid is refluxed back into the system (6), and the non-condensable gases tapped off.
Sufficiently heavy products flow back through the trays to the reactor bed, where they
are cracked further before revolatilising and entering the column again.
While such a system does offer a possible solution to the wide range of products
produced using microwave-assisted pyrolysis, it is also a significantly more complex
and costly design than the simple microwave reactors that have been used to study
microwave-assisted pyrolysis in the literature thus far, likely requiring significant scale
in order to operate economically. A comprehensive model would need to be developed
before commissioning such a scheme, not only of the energy and material inputs and
outputs, but also of the economics underlying the process. A capital-intense plant
negates one of the advantages of microwave-assisted pyrolysis, with its ability to
operate effectively and cheaply on a range of scales.
7.2 Further Study 191
7.2.6.2 Optimising for Gaseous Hydrocarbon Production
An alternative to the fractionating column reactor presented above is to construct a
reactor that is optimised to produce only non-condensable gases, but use the physical
properties of these as a separations system (as only the smallest hydrocarbons exist in
the gaseous state at room temperature). The mechanics of such an idea could also
operate with the reactor integrated into the condensing system. The fundamental
principle underlying this concept is for there to be a dramatic temperature gradient
between the hot reactor bed and the condensing column. The ability of microwave
heating to target only microwave-receptive materials provides a means of achieving
this, with input energy dispersed entirely within the carbon bed and a steep fall-off
in temperature as distance from the bed increases. By having a simple condenser
(effectively just sufficiently high reactor walls), the temperature gradient would be
such that liquids and aerosol particulates would fall back into the hot zone of the bed
to be re-cracked, and only short-chain gases could leave and be collected.
This method could be employed to obtain a consistent gaseous output stream
without catalysis, and indeed would maximise product yield by avoiding the formation
of char that is promoted with catalytic cracking. Any char that was produced (which
from the results presented in this study would be expected to be inconsequential) would
become part of the reactor bed.
A preliminary investigation into the feasibility of this concept showed positive
results. Using a small scale quartz reactor with a long vertical glass column that fed
into a series of Liebig condensers and a dry ice trap from the top of the column (a small
modification of the apparatus described in [150]), resulted in the formation of a clearly
distinct “inversion layer” of visible opaque pyrolysis aerosol/gas that did not reach the
top of the column. Almost no pyrolysis products were able to be condensed, with the
exiting products consisting almost entirely of short-chain hydrocarbons. While this one
brief experiment on a small-scale system demonstrates that such a system is potentially
viable, further work using similar equipment could develop this promising method fully.
In any case, the system described here compares favourably to other work attempt-
ing to achieve the same end; in an attempt to produce light gaseous olefins from
HDPE, Artetxe et al. [305] used a conventionally heated conical spouted bed reactor to
pyrolyse HDPE, producing long-chain hydrocarbons (waxes at room temperature) that
were passed through a cyclone to remove the bed material before entering a second
fixed bed HZSM-5 zeolite catalytic reactor that finally resulted in the production of
C2–C4 olefins. While an interesting and apparently successful scheme, the equivalent
192 Project Overview and Further Study
system proposed here is significantly simpler in operation (with the associated cost and
reliability implications), and does not involve the expense of a second catalytic reactor,
or the deactivation concerns (as seen in this work) that the use of a catalyst in such a
context implies.
7.2.7 Contamination and Diversification of Input Streams
As an exploratory study, this work has focused exclusively on the pyrolysis of HDPE
as a model plastic for simplicity’s sake. This qualification is not limited to this study
alone, with HDPE the only plastic in the literature that has been well characterised
with microwave-assisted pyrolysis. The development of a microwave-assisted pyrolysis
process able to cope with post-consumer waste would necessitate the ability to cope
with multiple materials simultaneously, as well as contamination of the sort found in
kerb-side collected plastics. There is a large potential to conduct further study in these
areas, with the end goal to develop a process sufficiently robust as to produce a high
value product from municipal solid waste.
7.2.8 Energy Return on Investment
An important consideration in the evaluation of the value of the microwave-assisted
pyrolysis process is the amount of energy made available in the final product relative
to the energy required to produce the product. The energy return on investment (EROI)
is calculated as follows:
EROI =
Energy of product
Energy required to produce product
Values of EROI below one indicate that more energy is consumed in the process than is
gained in the final product, while values above one give the multiple of energy gained
from the input of one unit. The products of the microwave-assisted pyrolysis of HDPE
have a calorific value, i.e., the energy that would be released if they were completely
combusted. Future study could determine this by the analysis of a representative
sample of the products with a bomb calorimeter. By comparing this to the electrical
energy required to operate the magnetron and other apparatus it would be possible to
calculate the EROI for the microwave-assisted pyrolysis process as performed in this
work.
Based on the work of Lam et al. [152], who conducted the microwave-assisted
pyrolysis of waste engine oil with similar apparatus, it is highly likely that the process
7.2 Further Study 193
described in this work would result in a positive EROI. Lam et al. observed an EROI of
up to 8 with their setup, using the methodology similar to that described above; this
compares with the EROI of crude oil produced in the USA of 11–18 in the early 2000s
(which has likely declined since then) [44]. However, the study carried out by Lam et al.
used a prototype reactor, with little attempt made to insulate the reactor and associated
fittings from heat loss, and with no attempt made to recover heat energy during the
cooling of products in the condensing system (a similar state of affairs was the case
with the apparatus used in this study). It is expected that considerable efficiency gains
(and a correspondingly higher EROI) would be made in the scale-up of the process,
the insulation of the reactor to avoid heat losses, and the recovery of heat from the
produced pyrolysis products.
It would also be interesting to conduct calorimetric analyses on the pyrolysis of
plastic using microwave heating and catalysts, and investigate how the use of various
catalysts is able to change the activation energy of the pyrolysis process, as well as the
potential efficiency gains arising from directly heating the catalyst itself.
7.2.9 Economic Analysis of the Process
A further important consideration in the evaluation of any future microwave-assisted
pyrolysis process for future development is whether or not the process would be
economically viable. Ideally, a comprehensive evaluation of all the cost and revenue
streams would be undertaken, including: the capital cost of any potential plant, the
electrical (and any other) running costs of the process, the cost of the (potentially
catalytic) reactor bed and any replacement of this that needs to occur, the cost of any
further separation or clean-up of the products required before sale, the market value
of the produced products and best route of sale for these, and any potential gate fee
that could be collected from disposing of the waste.
While it is not proposed that the apparatus described in this work be directly used to
create a commercial venture, the results contained in this study nevertheless provide
information that would be valuable if such an undertaking were carried out (e.g., the
comparison of batch vs. continuous operation modes, the composition of the various
product fractions with activated carbon catalysis and in the non-catalytic mode), and
would provide a foundation from which estimates of costs and revenues could based.
194 Project Overview and Further Study
7.3 Conclusions and Evaluation of Study Objectives
In light of the preceding overview, it is prudent to revisit the objectives of this study set
out at the beginning of this dissertation, and assess the progress made towards their
fulfilment. The following is a candid evaluation of the work carried out in regard to
each of these objectives in turn.
To develop suitable apparatus to reliably carry out the microwave-assisted
pyrolysis of HDPE, and collect the resulting pyrolysis products.
The inherited reactor described at the beginning of this dissertation possessed numer-
ous issues that prevented it from being used to conduct microwave-assisted pyrolysis
of HDPE. The development of the plunger injector and subsequent continuous rotary
feeder facilitated the introduction of the HDPE pellets into the reactor, while the
high levels of reflected power and arcing were tackled by a systematic approach to
identifying and eliminating failure points such as through simplification of the power
transmission system, and replacing the quartz window with a redesigned silicone
window that integrated the nitrogen gas inlet and water cooling. These changes,
combined with a computer-based control system that reacted to failures more rapidly
than a human could, dramatically enhanced reliability of the apparatus. After several
design iterations, the percolating dry ice condenser maximised collection of the
pyrolysis condensate, while the gas sampling system facilitated capture and analysis
of the remaining pyrolysis products. The combination of these components saw a
functioning microwave-assisted pyrolysis reactor that performed reliably throughout
the remainder of the work described in this dissertation.
To determine whether or not activated carbon is able to act as a catalyst in the
microwave-assisted pyrolysis of HDPE.
The dramatic difference first seen between the products produced using activated
carbon and the non-catalytic coke bed conclusively demonstrated that virgin activated
carbon is an effective catalyst in the microwave-assisted pyrolysis of HDPE. The
resulting pyrolysis products produced in this manner underwent significantly more
cracking, resulting in a product containing a much narrower range of hydrocarbons.
Unfortunately, activated carbon deactivated relatively rapidly when used as a catalyst
for this purpose, with the catalytic activity and resulting effect on produced products
declining over time.
7.3 Conclusions and Evaluation of Study Objectives 195
To characterise the products produced by the microwave-assisted pyrolysis of
HDPE, and evaluate their potential utility.
GC/MS analysis proved an effective tool to characterise the products produced from
the microwave-assisted pyrolysis of HDPE. With no catalyst, the pyrolysis products
comprised almost entirely of linear aliphatics ranging in size from C1–C55. Using a
virgin activated carbon bed the products ranged from C1–C35 across all temperatures
studied, though the majority were less than this maximum, with more than 98% of the
condensate produced at 450°C in the liquid transport fuel range of C5–C21. Indeed,
the condensate produced using virgin activated carbon could be used as a transport
fuel relatively easily, though the aromatic fraction may have to be separated in order to
comply with maximum limits for these compounds in these applications. This aromatic
fraction is highly valuable in and of itself, with uses such as solvents, and as precursors
for a wide variety of compounds including drugs, lubricants, detergents, plastics (e.g.,
polystyrene, polycarbonate), and explosives. The condensate produced using the non-
catalytic coke bed contained a substantial portion of high molecular weight waxes that
would need to be separated out before this fraction could be put to use as a liquid
fuel, and which have low value themselves. The non-condensable gases in all cases
were short-chain aliphatics which could be used as fuels (e.g., natural gas, LPG) or as
feedstock for a range of chemicals (ethene, for example, being the most widely used
organic compound in the chemical industry, and is the monomer for HDPE). As the
activated carbon deactivated, the profile of the produced products moved toward those
produced using the coke bed, with a corresponding reduction in more useful short-
chain hydrocarbons and an increase in the less useful waxes.
To examine the scale-up of the process with a larger capacity and continuous
throughput.
The design and fabrication of the continuous rotary feeder successfully allowed the
continuous addition of HDPE with an associated expansion in processing capacity.
Through the processing of a larger quantity of HDPE it became clear that the activated
carbon deactivated relatively rapidly, presenting a substantial difficulty in the potential
scale-up of the process owing to the high cost of replacing the catalyst after its
deactivation.
196 Project Overview and Further Study
To understand and characterise the catalytic properties of activated carbon in
the context of the pyrolytic breakdown of HDPE.
The mechanisms presented in chapter 6 are able to account for the species observed
in the pyrolysis products in both the catalytic and non-catalytic case. Specifically,
activated carbon has a very high surface area that contains high energy active sites
that facilitate electron and hydrogen transfer, reducing the activation energy of radical-
mediated cracking mechanisms. For example, the reduction in activation energy
of aromatic formation allows these more stable hydrocarbons to form, in greater
quantities and at a lower temperature. The integration and dehydrogenation of
hydrocarbon onto active sites deactivates them and results in the deposition of a non-
catalytic low surface area carbon that reduces the effectivity of the catalyst.
To evaluate the potential of microwave-assisted pyrolysis of HDPE for future
commercial development on an industrial scale.
Microwave-assisted pyrolysis of HDPE using the method as described in this work
is not likely to be feasible because of the high cost of the activated carbon, which
would need to be frequently replaced. A detailed economic analysis of the process
was not conducted owing to the need for a solution to this issue before the prospect
of a commercially viable process could be realised. Accordingly, several possibilities
were presented that could provide the focus of future work towards this goal, including
the reactivation of the activated carbon, either in-situ or in an online regenerator, the
incorporation of a fractional separator into the reactor in order to separate the pyrolysis
into multiple output streams, and the design of a new reactor built with the intention
to exclusively produce short-chain non-condensable gases — this would eliminate the
need to separate fractions from one another, and could be designed to be operated
with either a catalytic or non-catalytic bed. Given the high aromatic content of the
products produced using activated carbon, it is likely that future development of a
novel process utilising activated carbon as a catalyst would focus on the production of
these compounds.
7.3.1 Additional Conclusions
In addition to meeting the objectives laid out at the beginning of this study, several
further points worthy of discussion came to light as a result of the intervening
work conducted. In general terms, this study contributed to a more complete
7.4 Closing Remarks 197
characterisation of the products of the microwave-assisted pyrolysis of HDPE, and
the reaction parameters under which they were produced. In particular, this study
highlighted the importance of residence time in determining the profile of the produced
products and the interplay this parameter has with the reaction temperature. This
study was also able to provide information on the influence of the mode of sample
introduction (i.e., batch or continuous), and how this affects residence time and product
composition via reaction kinetics. Characterising the influence of these parameters
could help to tailor future processes to target particular products (e.g., enhancing
aromatic formation by engineering a longer residence time).
This work also presents the most complete account so far of the use of activ-
ated carbon to pyrolyse long-chain hydrocarbons, with the catalytic and deactivation
mechanisms married to the carbon’s structural and physical properties (e.g., porosity
correlating to catalytic ability), with these mechanisms and the product profile (e.g.,
lack of branched compounds) suggesting a different catalytic mechanism to that of
traditional cracking acid catalysts. Aside from its rapid deactivation, which may or may
not pose a surmountable issue in future reactor designs, the knowledge of the output
products, including the ability to produce a very high proportion of aromatics, even at
low reaction temperatures, provides a potential opportunity for future researchers to
capitalise on, while the possibilities described in the future work section of this chapter
indicate potential paths towards several other avenues of future exploration.
7.4 Closing Remarks
As with any doctoral thesis, the ultimate goal of this study was to make a contribution
to the total sum of human knowledge. The development of any chemical process
necessarily starts out small, and proceeds in incremental steps, each of which introduce
new complexities and issues that must be overcome. While the deactivation of activated
carbon prevents its use as a catalyst in the pyrolysis of long-chain hydrocarbons,
at least in the configuration used in this work, it cannot be argued that this study
did not produce valuable knowledge. Not only has this study provided an important
contribution to the understanding of the use of activated carbon as a cracking catalyst,
using the microwave-assisted pyrolysis of HDPE as an informative case study, but the
knowledge from this study has also indicated where the future prospects of activated
carbon catalysis may lie. It is hoped that out of the knowledge contained in this study,
a future economically beneficial process will be developed that will help to maximise
198 Project Overview and Further Study
the utilisation of hydrocarbons such as plastics, and contribute to an improvement in
the efficiency with which humanity uses its increasingly scarce and difficult to obtain
resources.
References
[1] Plastics Europe. Plastics — the Facts. An analysis of European plastics
production, demand and recovery. Technical report, Association of Plastics
Manufacturers, Brussels, 2011.
[2] Plastics Challenge 2020. Founding Commitments.
http://plastics2020challenge.com, Accessed 22/08/2012, 2010.
[3] Scott G. Polymers and the Environment. Royal Society of Chemistry, Cambridge,
1999.
[4] Sorum L., Gronli M.G., and Hustad J.E. Pyrolysis characteristics and kinetics of
municipal solid wastes. Fuel, 80: 1217–1227, 2001.
[5] Rowat S.C. Incinerator toxic emissions: a brief summary of human health effects
with a note on regulatory control. Medical hypotheses, 52: 389–396, 1999.
[6] Cormier S.A., Lomnicki S., Backes W., and Dellinger B. Origin and Health
Impacts of Emissions of Toxic By-Products and Fine Particles from Combustion
and Thermal Treatment of Hazardous Wastes and Materials. Environmental
Health Perspectives, 114: 810–817, 2006.
[7] Adriano D.C., Page A.L., Elseewi A.A., Chang A.C., and Straughan I. Utilization
and Disposal of Fly Ash and Other Coal Residues in Terrestrial Ecosystems: A
Review, 1980.
[8] Dry M.E. The Fischer-Tropsch process: 1950-2000. Catalysis Today, 71: 227–
241, 2002.
[9] Dry M.E. High quality diesel via the Fischer-Tropsch process — a review. Journal
of Chemical Technology & Biotechnology, 77: 43–50, 2002.
[10] Aguado J., Serrano D.P., and San Miguel G. European Trends in the Feedstock
Recycling of Plastic Wastes. Global NEST Journal, 9: 12–19, 2007.
[11] Mastral F., Esperanza E., Garcia P., and Juste M. Pyrolysis of high-density
polyethylene in a fluidised bed reactor. Influence of the temperature and
residence time. Journal of Analytical and Applied Pyrolysis, 63: 1–15, 2002.
[12] Domínguez A., Menéndez J.A., Fernández Y., Pis J.J., Nabais J.M.V., Carrott P.J.M.,
and Carrott M.M.L.R. Conventional and microwave induced pyrolysis of coffee
hulls for the production of a hydrogen rich fuel gas. Journal of Analytical and
Applied Pyrolysis, 79: 128–135, 2007.
[13] Buekens A. and Huang H. Catalytic plastics cracking for recovery of gasoline-
range hydrocarbons from municipal plastic wastes. Resources, Conservation and
Recycling, 23: 163–181, 1998.
[14] Kumar S., Panda A.K., and Singh R. A review on tertiary recycling of high-density
polyethylene to fuel. Resources, Conservation and Recycling, 55: 893–910, 2011.
[15] Robinson J., Snape C., Kingman S., and Shang H. Thermal desorption and
pyrolysis of oil contaminated drill cuttings by microwave heating. Journal of
Analytical and Applied Pyrolysis, 81: 27–32, 2008.
199
200 REFERENCES
[16] Jones D.A., Lelyveld T.P., Mavrofidis S.D., Kingman S.W., and Miles N.J.
Microwave heating applications in environmental engineering — a review.
Resources, Conservation and Recycling, 34: 75–90, 2002.
[17] Marken F., Sur U.K., Coles B.A., and Compton R.G. Focused microwaves in
electrochemical processes. Electrochimica Acta, 51: 2195–2203, 2006.
[18] Zhang X. and Hayward D.O. Applications of microwave dielectric heating in
environment-related heterogeneous gas-phase catalytic systems. Inorganica
Chimica Acta, 359: 3421–3433, 2006.
[19] Miura M., Kaga H., Tanaka S., Takahashi K., and Ando K. Rapid Microwave
Pyrolysis of Wood. Journal of Chemical Engineering of Japan, 33: 299–302, 2000.
[20] Lei H., Ren S., and Julson J. The Effects of Reaction Temperature and Time and
Particle Size of Corn Stover on Microwave Pyrolysis. Energy & Fuels, 23: 3254–
3261, 2009.
[21] Menéndez J.A., Dominguez A., Inguanzo M., and Pis J.J. Microwave pyrolysis of
sewage sludge: analysis of the gas fraction. Journal of Analytical and Applied
Pyrolysis, 71: 657–667, 2004.
[22] Fernández Y., Arenillas A., Bermúdez J., and Menéndez J. Comparative study
of conventional and microwave-assisted pyrolysis, steam and dry reforming
of glycerol for syngas production, using a carbonaceous catalyst. Journal of
Analytical and Applied Pyrolysis, 88: 155–159, 2010.
[23] Udalov E.I., Bolotov V.A., Tanashev Y.Y., Chernousov Y.D., and Parmon V.N.
Pyrolysis of liquid hexadecane with selective microwave heating of the catalyst.
Theoretical and Experimental Chemistry, 46: 370–377, 2011.
[24] Zuo W., Tian Y., and Ren N. The important role of microwave receptors in
bio-fuel production by microwave-induced pyrolysis of sewage sludge. Waste
management, 31: 1321–1326, 2011.
[25] Haque K.E. Microwave energy for mineral treatment processes - a brief review.
International Journal of Mineral Processing, 57: 1–24, 1999.
[26] Osepchuk J.M. and Fellow L. Microwave Power Applications. IEEE Transactions
on Microwave Theory and Techniques, 50: 975–985, 2002.
[27] Holland K.M. Process of destructive distillation of organic material, United States
Patent 5330623, 1994.
[28] Holland K.M. Apparatus for waste pyrolysis, United States Patent 5387321, 1995.
[29] Ludlow-Palafox C. and Chase H.A. Microwave-Induced Pyrolysis of Plastic
Wastes. Industrial & Engineering Chemistry Research, 40: 4749–4756, 2001.
[30] Undri A., Rosi L., and Frediani M. Microwave Pyrolysis of Polymeric Materials.
InTech, 2011.
[31] Anthony E. Fluidized bed combustion of alternative solid fuels; status, successes
and problems of the technology. Progress in Energy and Combustion Science,
21: 239–268, 1995.
[32] Anthony E.J., Iribarne A.P., Iribarne J.V., Talbot R., Jia L., and Granatstein D.L.
Fouling in a 160 MWe FBC boiler firing coal and petroleum coke. Fuel, 80: 1009–
1014, 2001.
[33] Bryers R.W. Utilization of petroleum coke and petroleum coke/coal blends as a
means of steam raising. Fuel Processing Technology, 44: 121–141, 1995.
[34] Juntgen H. Activated carbon as catalyst support: A review of new research
results. Fuel, 65: 1436–1446, 1986.
REFERENCES 201
[35] Aksoylu A., Madalena M., Freitas a., Pereira M.R., and Figueiredo J.L. The
effects of different activated carbon supports and support modifications on the
properties of Pt/AC catalysts. Carbon, 39: 175–185, 2001.
[36] Rodriguez-Reinoso F. The role of carbon materials in heterogeneous catalysis.
Carbon, 36: 159–175, 1998.
[37] Trimm D.L. Catalysis, Vol. 4. The Royal Society of Chemistry, London, U.K., 1981.
[38] Pereira M.F.R., Orfao J.J.M., and Figueiredo J.L. Oxidative dehydrogenation of
ethylbenzene on activated carbon catalysts. I. Influence of surface chemical
groups. Applied Catalysis A: General, 184: 153–160, 1999.
[39] Muradov N., Smith F., and T-Raissi A. Catalytic activity of carbons for methane
decomposition reaction. Catalysis Today, 102-103: 225–233, 2005.
[40] Key World Energy Statistics. Technical report, International Energy Agency,
Paris, 2010.
[41] Klein D. Organic Chemistry. John Wiley and Sons, 2001.
[42] Nadim F., Liu S., Hoag G., and Chen J. A comparison of spectrophotometric and
gas chromatographic measurements of heavy petroleum products in soil samples.
Water, Air, and Soil Pollution, 134: 97–109, 2002.
[43] Anderson R. and Kahya D. The cost of petrol and oil: How it breaks down, 2011.
[44] Hall C.A.S., Balogh S., and Murphy D.J. What is the Minimum EROI that a
Sustainable Society Must Have? Energies, 2: 25–47, 2009.
[45] Cossee P. Ziegler-Natta catalysis I. Mechanism of polymerization of [alpha]-
olefins with Ziegler-Natta catalysts. Journal of Catalysis, 3: 80, 1964.
[46] Ludlow-Palafox C. Microwave Induced Pyrolysis of Plastic Wastes. PhD Thesis,
University of Cambridge, 2001.
[47] Gu Xiasheng H.S.K. The Status and Trend of Water Pollution Control Technology
in China. Water International, 7: 78–80, 1982.
[48] Linkov I. and Bridges T.S. (Eds.). Climate. NATO Science for Peace and Security
Series C: Environmental Security. Springer Netherlands, Dordrecht, 2011.
[49] Williams A.T. and Simmons S.L. Sources of riverine litter: the river Taff, South
Wales, UK. Water, Air, and Soil Pollution, 112: 197–216, 1999.
[50] Lodders K. Solar System Abundances and Condensation Temperature of the
Elements. The Astrophysics Journal, 591: 1220–1247, 2003.
[51] Strock M. Carbon Allotropes. http://commons.wikimedia.org/wiki/
File:Eight_Allotropes_of_Carbon.png, 2006.
[52] Marsh H. and Rodriguez-Reinoso F. Activated Carbon. Elsevier Ltd, Burlington,
2006.
[53] Franklin R.E. The interpretation of diffuse X-ray diagrams of carbon. Acta
Crystallographica, 3: 107–121, 1950.
[54] Franklin R.E. The structure of graphitic carbons. Acta Crystallographica, 4:
253–261, 1951.
[55] Benaddi H., Bandosz T., Jagiello J., Schwarz J., Rouzaud J., Legras D., and Béguin
F. Surface functionality and porosity of activated carbons obtained from chemical
activation of wood. Carbon, 38: 669–674, 2000.
[56] Srinivasakannan C. Production of activated carbon from rubber wood sawdust.
Biomass and Bioenergy, 27: 89–96, 2004.
202 REFERENCES
[57] Khezami L., Chetouani A., Taouk B., and Capart R. Production and
characterisation of activated carbon from wood components in powder:
Cellulose, lignin, xylan. Powder Technology, 157: 48–56, 2005.
[58] Kirubakaran C.J., Krishnaiah K., and Seshadrit S.K. Experimental Study of the
Production of Activated Carbon from Coconut Shells in a Fluidized Bed Reactor.
Industrial & Engineering Chemistry Research, 30: 2411–2416, 1991.
[59] Gratuito M.K.B., Panyathanmaporn T., Chumnanklang R.a., Sirinuntawittaya N.,
and Dutta a. Production of activated carbon from coconut shell: optimization
using response surface methodology. Bioresource technology, 99: 4887–4895,
2008.
[60] Satya Sai P.M., Ahmed J., and Krishnaiah K. Production of Activated Carbon
from Coconut Shell Char in a Fluidized Bed Reactor. Industrial & Engineering
Chemistry Research, 36: 3625–3630, 1997.
[61] Aygun A., Yenisoy-Karakas S., and Duman I. Production of granular activated
carbon from fruit stones and nutshells and evaluation of their physical, chemical
and adsorption properties. Microporous and Mesoporous Materials, 66: 189–
195, 2003.
[62] Lafi W.K. Production of activated carbon from acorns and olive seeds. Biomass
and Bioenergy, 20: 57–62, 2001.
[63] Haimour N.M. and Emeish S. Utilization of date stones for production of activated
carbon using phosphoric acid. Waste management (New York, N.Y.), 26: 651–660,
2006.
[64] Kim J.W., Sohn M.H., Kim D.S., Sohn S.M., and Kwon Y.S. Production of granular
activated carbon from waste walnut shell and its adsorption characteristics for
Cu2+ ion. Journal of hazardous materials, 85: 301–315, 2001.
[65] Ahmadpour A. and Do D.D. The preparation of activated carbon from macadamia
nutshell by chemical activation. Carbon, 35: 1723–1732, 1997.
[66] Teng H. and Lin H.c. Activated carbon production from low ash subbituminous
coal with CO2 activation. AIChE Journal, 44: 1170–1177, 1998.
[67] Jüntgen H., Knoblauch K., and Harder K. Carbon molecular sieves: production
from coal and application in gas separation. Fuel, 60: 817–822, 1981.
[68] Lozano-Castello D., Lillo-Ródenas M.A., Cazorla-Amorós D., and Linares-Solano
A. Preparation of activated carbons from Spanish anthracite: I. Activation by
KOH. Carbon, 39: 741–749, 2001.
[69] DiPanfilo R. and Egiebor N. Activated carbon production from synthetic crude
coke. Fuel processing technology, 46: 157–169, 1996.
[70] Gonzalez J., Gonzalez M., Molina-Sabio M., Rodriguez-Reinoso F., and Sepulveda-
Escribano A. Porosity of activated carbons prepared from different lignocellulosic
materials. Carbon, 338: 1175–1177, 1995.
[71] Rodriguez-Reinoso F. and Molina-Sabio M. Textural and chemical characteriza-
tion of microporous carbons. Advances in Colloid and Interface Science, 76-77:
271–294, 1998.
[72] Caturla F., Molina-Sabio M., and Rodriguez-Reinoso F. Preparation of activated
carbon by chemical activation with ZnCl2. Carbon, 29: 999–1007, 1991.
[73] Molina-Sabio M. and Rodriguez-Reinoso F. Role of chemical activation in the
development of carbon porosity. Colloids and Surfaces A: Physicochemical and
Engineering Aspects, 241: 15–25, 2004.
REFERENCES 203
[74] Hayashi J., Kazehaya A., Muroyama K., and Watkinson A. Preparation of activated
carbon from lignin by chemical activation. Carbon, 38: 1873–1878, 2000.
[75] Marsh H., Yan D., O’Grady T., and Wennerberg A. Formation of active carbons
from cokes using potassium hydroxide. Carbon, 22: 603–611, 1984.
[76] Shih T.C., Wangpaichitr M., and Suffet M. Evaluation of granular activated
carbon technology for the removal of methyl tertiary butyl ether (MTBE) from
drinking water. Water research, 37: 375–385, 2003.
[77] Rengaraj S., Moon S.H., Sivabalan R., Arabindoo B., and Murugesan V.
Agricultural solid waste for the removal of organics: adsorption of phenol from
water and wastewater by palm seed coat activated carbon. Waste management
(New York, N.Y.), 22: 543–548, 2002.
[78] Giusti D., Conway R., and Lawson C. Activated carbon adsorption of
petrochemicals. Journal (Water Pollution Control . . . , 46: 947–965, 1974.
[79] Ternes T.a., Meisenheimer M., McDowell D., Sacher F., Brauch H.J., Haist-Gulde
B., Preuss G., Wilme U., and Zulei-Seibert N. Removal of pharmaceuticals during
drinking water treatment. Environmental science & technology, 36: 3855–3863,
2002.
[80] Yoon Y., Westerhoff P., Snyder S.A., and Esparza M. HPLC-fluorescence detection
and adsorption of bisphenol A, 17beta-estradiol, and 17alpha-ethynyl estradiol
on powdered activated carbon. Water research, 37: 3530–3537, 2003.
[81] Snyder S.A., Adham S., Redding A.M., Cannon F.S., DeCarolis J., Oppenheimer J.,
Wert E.C., and Yoon Y. Role of membranes and activated carbon in the removal
of endocrine disruptors and pharmaceuticals. Desalination, 202: 156–181, 2007.
[82] Ebie K., Li F., Azuma Y., Yuasa A., and Hagishita T. Pore distribution effect of
activated carbon in adsorbing organic micropollutants from natural water. Water
research, 35: 167–179, 2001.
[83] Pelekani C. and Snoeyink V. Competitive adsorption in natural water: role of
activated carbon pore size. Water Research, 33: 1209–1219, 1999.
[84] Bjelopavlic M., Newcombe G., and Hayes R. Adsorption of NOM onto Activated
Carbon: Effect of Surface Charge, Ionic Strength, and Pore Volume Distribution.
Journal of colloid and interface science, 210: 271–280, 1999.
[85] Mohan D. and Singh K.P. Single- and multi-component adsorption of cadmium
and zinc using activated carbon derived from bagasse-an agricultural waste.
Water research, 36: 2304–2318, 2002.
[86] Huang C. and Fu P. Treatment of arsenic (V)-containing water by the activated
carbon process. Journal (Water Pollution Control Federation), 56: 233–242, 1984.
[87] Kobya M., Demirbas E., Senturk E., and Ince M. Adsorption of heavy metal
ions from aqueous solutions by activated carbon prepared from apricot stone.
Bioresource technology, 96: 1518–21, 2005.
[88] Leggott N.L., Shephard G.S., Stockenström S., Staal E., and van Schatkwyk
D.J. The reduction of patulin in apple juice by three different types of activated
carbon. Food additives and contaminants, 18: 825–9, 2001.
[89] Ahmedna M., Johns M.M., Clarke S.J., Marshall W.E., and Rao R.M. Potential
of agricultural by-product-based activated carbons for use in raw sugar
decolourisation. Journal of the Science of Food and Agriculture, 75: 117–124,
1997.
[90] Ahmedna M., Marshall W., and Rao R. Surface properties of granular
activated carbons from agricultural by-products and their effects on raw sugar
decolorization. Bioresource Technology, 71: 103–112, 2000.
204 REFERENCES
[91] Pandey R., Sanyal P., Chattopadhyay N., and Kaul S. Treatment and reuse of
wastes of a vegetable oil refinery. Resources, Conservation and Recycling, 37:
101–117, 2003.
[92] Dutta M., Dutta N., and Bhattacharya K. Aqueous phase adsorption of certain
beta-lactam antibiotics onto polymeric resins and activated carbon. Separation
and Purification Technology, 16: 213–224, 1999.
[93] Tamai H., Kakii T., Hirota Y., Kumamoto T., and Yasuda H. Synthesis of
Extremely Large Mesoporous Activated Carbon and Its Unique Adsorption for
Giant Molecules. Chemistry of Materials, 8: 454–462, 1996.
[94] Ruhl M. Recovery of VOCs via adsorption on activated carbon. Chemical
Engineering Progress, pp. 37–41, 1993.
[95] Gurrath M., Kuretzky T., and Boehm H. Palladium catalysts on activated
carbon supports: Influence of reduction temperature, origin of the support and
pretreatments of the carbon surface. Carbon, 38: 1241–1255, 2000.
[96] Yang Y., Chiang K., and Burke N. Porous carbon-supported catalysts for energy
and environmental applications: A short review. Catalysis Today, 178: 197–205,
2011.
[97] Prado-Burguete C., Linares-Solano A., Rodríguez-Reinoso F., and de Lecea C.M.
The effect of oxygen surface groups of the support on platinum dispersion in
Pt/carbon catalysts. Journal of Catalysis, 115: 98–106, 1989.
[98] Suh D.J., Park T.J., and Ihm S.K. Characteristics of carbon-supported palladium
catalysts for liquid-phase hydrogenation of nitroaromatics. Industrial &
Engineering Chemistry Research, 31: 1849–1856, 1992.
[99] Auer E., Freund a., Pietsch J., and Tacke T. Carbons as supports for industrial
precious metal catalysts. Applied Catalysis A: General, 173: 259–271, 1998.
[100] Ma W., Ding Y., and Lin L. Fischer-Tropsch synthesis over activated-carbon-
supported cobalt catalysts: Effect of Co loading and promoters on catalyst
performance. Industrial & engineering chemistry . . . , 43: 2391–2398, 2004.
[101] Ma W., Kugler E.L., Wright J., and Dadyburjor D.B. Mo-Fe catalysts supported
on activated carbon for synthesis of liquid fuels by the Fischer-Tropsch process:
effect of Mo addition on reducibility, activity, and hydrocarbon. Energy & Fuels,
20: 2299–2307, 2006.
[102] Ma W., Kugler E.L., and Dadyburjor D.B. Effect of Properties of Various Activated-
Carbon Supports and Supported Fe-Mo-Cu-K Catalysts on Metal Precursor
Distribution, Metal Reduction, and Fischer-Tropsch Synthesis. Energy & Fuels,
24: 4099–4110, 2010.
[103] Kubickova I., Snare M., Eranen K., Maki-Arvela P., and Murzin D.Y. Hydrocarbons
for diesel fuel via decarboxylation of vegetable oils. Catalysis Today, 106: 197–
200, 2005.
[104] Baroutian S., Aroua M.K., Raman A.A.A., and Sulaiman N.M.N. A packed bed
membrane reactor for production of biodiesel using activated carbon supported
catalyst. Bioresource technology, 102: 1095–1102, 2011.
[105] Dalai A.K., Majumdar a., and Tollefson E.L. Low Temperature Catalytic Oxidation
of Hydrogen Sulfide in Sour Produced Wastewater Using Activated Carbon
Catalysts. Environmental Science & Technology, 33: 2241–2246, 1999.
[106] Wu X., Kercher A.K., Schwartz V., Overbury S.H., and Armstrong T.R. Activated
carbons for selective catalytic oxidation of hydrogen sulfide to sulfur. Carbon,
43: 1087–1090, 2005.
REFERENCES 205
[107] Richter E. Carbon catalysts for pollution control. Catalysis Today, 7: 93–112,
1990.
[108] de Jesús Díaz Velásquez J., Suárez L.M.C., and Figueiredo J.L. Oxidative
dehydrogenation of isobutane over activated carbon catalysts. Applied Catalysis
A: General, 311: 51–57, 2006.
[109] Lisovskii A. and Aharoni C. Carbonaceous deposits as catalysts for
oxydehydrogenation of alkylbenzenes. Catalysis Reviews: Science and
Engineering, 36: 25–74, 1994.
[110] Pereira M., Orfao J., and Figueiredo J. Oxidative dehydrogenation of ethylbenzene
on activated carbon catalysts: 2. Kinetic modelling. Applied Catalysis A: General,
196: 43–54, 2000.
[111] Pereira M., Orfao J., and Figueiredo J. Oxidative dehydrogenation of ethylbenzene
on activated carbon catalysts: 3. Catalyst deactivation. Applied Catalysis A:
General, 218: 307–318, 2001.
[112] Grunewald G. and Drago R. Oxidative dehydrogenation of ethylbenzene to
stykene over carbon-based catalysts. Journal of Molecular Catalysis, 58: 227–
233, 1990.
[113] Guerrero-Ruiz a. and Rodríguez-Ramos I. Oxydehydrogenation of ethylbenzene
to styrene catalyzed by graphites and activated carbons. Carbon, 32: 23–29,
1994.
[114] Szymanski G.S. and Rychlicki G. Importance of oxygen surface groups in catalytic
dehydration and dehydrogenation of butan-2-ol promoted by carbon catalysts.
Carbon, 29: 489–498, 1991.
[115] Szymanski G. and Rychlicki G. Catalytic conversion of propan-2-ol on carbon
catalysts. Carbon, 31: 247–257, 1993.
[116] Szymanski G., Rychlicki G., and Terzyk A. Catalytic conversion of ethanol on
carbon catalysts. Carbon, 32, 1994.
[117] Muradov N., Smith F., Huang C., and T-Raissi A. Autothermal catalytic pyrolysis
of methane as a new route to hydrogen production with reduced CO2 emissions.
Catalysis Today, 116: 281–288, 2006.
[118] Muradov N.Z. How to produce hydrogen from fossil fuels without CO2 emission.
International Journal of Hydrogen Energy, 18: 211–215, 1993.
[119] Muradov N.Z. CO2-Free Production of Hydrogen by Catalytic Pyrolysis of
Hydrocarbon Fuel. Energy & Fuels, 12: 41–48, 1998.
[120] Muradov N. Catalysis of methane decomposition over elemental carbon.
Catalysis Communications, 2: 89–94, 2001.
[121] Muradov N. Hydrogen via methane decomposition: an application for
decarbonization of fossil fuels. International Journal of Hydrogen Energy, 26:
1165–1175, 2001.
[122] Jung J., Nam W., Yoon K., and Han G. Hydrogen production by catalytic
decomposition of methane over carbon catalysts in a fluidized bed. Korean
Journal of Chemical . . . , 24: 674–678, 2007.
[123] Petkovic L.M., Ginosar D.M., Rollins H.W., Burch K.C., Deiana C., Silva
H.S., Sardella M.F., and Granados D. Activated carbon catalysts for the
production of hydrogen via the sulfur-iodine thermochemical water splitting
cycle. International Journal of Hydrogen Energy, 34: 4057–4064, 2009.
[124] Meredith R. Engineers’ handbook of industrial microwave heating. The
Institution of Electrical Engineers, London, 1998.
206 REFERENCES
[125] Metaxas A. and Meredith R. Industrial Microwave Heating. Peter Peregrinus,
London, 1993.
[126] Metaxas A. Foundations of electroheat: a unified approach. John Wiley and Sons,
1996.
[127] Kaye and Laby. Tables of Physical & Chemical Constants. The National Physical
Laboratory, Teddington, 2006.
[128] Menéndez J.A., Arenillas A., Fidalgo B., Fernández Y., Zubizarreta L., Calvo E.,
and Bermúdez J. Microwave heating processes involving carbon materials. Fuel
Processing Technology, 91: 1–8, 2010.
[129] Atwater J. and Wheeler Jr. R. Microwave permittivity and dielectric relaxation
of a high surface area activated carbon. Applied Physics A: Materials Science &
Processing, 79: 125–129, 2004.
[130] Thostenson E. and Chou T.W. Microwave processing: fundamentals and
applications. Composites Part A: Applied Science and Manufacturing, 30: 1055–
1071, 1999.
[131] Fukushimaa H., Yamanakaa T., and Matsui M. Microwave heating of ceramics
and its application to joining. Journal of Materials Research, 5: 397–405, 1990.
[132] Marand E., Baker K.R., and Graybeal J.D. Reaction mechanisms of epoxy
resins undergoing thermal and microwave cure from in situ measurements of
microwave dielectric properties and infrared spectroscopy. Macromolecules, 25:
2243–2252, 1992.
[133] Lidström P., Tierney J., Wathey B., and Westman J. Microwave assisted organic
synthesis–a review. Tetrahedron, 57: 9225–9283, 2001.
[134] Gedye R.N., Smith F.E., and Westaway K.C. The rapid synthesis of organic
compounds in microwave ovens. Canadian Journal of Chemistry, 66: 17–26,
1988.
[135] Roberts B.A. and Strauss C.R. Toward rapid, "green", predictable microwave-
assisted synthesis. Accounts of chemical research, 38: 653–661, 2005.
[136] Wei J., Hawley M.C., and Demeuse M.T. Kinetics modeling and time-temperature-
transformation diagram of microwave and thermal cure of epoxy resins. Polymer
Engineering and Science, 35: 461–470, 1995.
[137] Clark D.E. and Sutton W.H. Microwave processing of materials. Annual Review
of Materials Science, 26: 299–331, 1996.
[138] de la Hoz A., Díaz-Ortiz A., and Moreno A. Microwaves in organic synthesis.
Thermal and non-thermal microwave effects. Chemical Society Reviews, 34: 164–
178, 2005.
[139] Hosseini M., Stiasni N., Barbieri V., and Kappe C.O. Microwave-assisted
asymmetric organocatalysis. A probe for nonthermal microwave effects and the
concept of simultaneous cooling. The Journal of organic chemistry, 72: 1417–24,
2007.
[140] Herrero M.A., Kremsner J.M., and Kappe C.O. Nonthermal microwave effects
revisited: on the importance of internal temperature monitoring and agitation in
microwave chemistry. The Journal of organic chemistry, 73: 36–47, 2008.
[141] Schmink J.R. and Leadbeater N.E. Probing "microwave effects" using Raman
spectroscopy. Organic & biomolecular chemistry, 7: 3842–6, 2009.
[142] Domínguez A., Fidalgo B., Fernandez Y., Pis J., and Menendez J. Microwave-
assisted catalytic decomposition of methane over activated carbon for CO2-free
hydrogen production. International Journal of Hydrogen Energy, 32: 4792–4799,
2007.
REFERENCES 207
[143] Muradov N.Z. The generation of microwave-induced plasma in granular active
carbons under fluidised bed conditions. Carbon, 46: 220–228, 2008.
[144] Fernández Y., Fidalgo B., Domínguez A., Arenillas A., and Menéndez J. Carbon
nanofilament synthesis by the decomposition of CH4/CO2 under microwave
heating. Carbon, 45: 1706–1709, 2007.
[145] Jiang S. Microwave pyrolysis of waste automotive tyres. CPGS Project Report,
University of Cambridge, 2004.
[146] Yatsun a.V., Konovalov P.N., and Konovalov N.P. Gaseous products of microwave
pyrolysis of scrap tires. Solid Fuel Chemistry, 42: 187–191, 2008.
[147] Lee A. Waving aside tyre incineration. Engineer, 292: 12, 2003.
[148] Sharma V., Fortuna F., Mincarini M., Berillo M., and Cornacchia G. Disposal
of waste tyres for energy recovery and safe environment. Applied Energy, 65:
381–394, 2000.
[149] Lam S.S., Russell A.D., and Chase H.A. Microwave pyrolysis, a novel process for
recycling waste automotive engine oil. Energy, 35: 2985–2991, 2010.
[150] Lam S.S., Russell A.D., and Chase H.A. Pyrolysis Using Microwave Heating: A
Sustainable Process for Recycling Used Car Engine Oil. Industrial & Engineering
Chemistry Research, 49: 10845–10851, 2010.
[151] Lam S.S. Microwave-induced Pyrolysis of Waste Automotive Oil. PhD Thesis,
University of Cambridge, 2011.
[152] Lam S.S., Russell A.D., Lee C.L., and Chase H.A. Microwave-heated pyrolysis
of waste automotive engine oil: Influence of operation parameters on the yield,
composition, and fuel properties of pyrolysis oil. Fuel, 92: 327–339, 2012.
[153] Menéndez J.a., Inguanzo M., and Pis J.J. Microwave-induced pyrolysis of sewage
sludge. Water research, 36: 3261–4, 2002.
[154] Menéndez J.A., Domínguez A., Inguanzo M., and Pis J.J. Microwave-induced
drying, pyrolysis and gasification (MWDPG) of sewage sludge: Vitrification of
the solid residue. Journal of Analytical and Applied Pyrolysis, 74: 406–412, 2005.
[155] Domínguez A., Menéndez J.A., Inguanzo M., and Pís J.J. Production of bio-fuels by
high temperature pyrolysis of sewage sludge using conventional and microwave
heating. Bioresource technology, 97: 1185–93, 2006.
[156] Domínguez A., Menéndez J.A., Inguanzo M., Bernad P.L., and Pis J.J. Gas
chromatographic-mass spectrometric study of the oil fractions produced
by microwave-assisted pyrolysis of different sewage sludges. Journal of
chromatography. A, 1012: 193–206, 2003.
[157] Domínguez a., Menéndez J.A., Inguanzo M., and Pis J.J. Investigations into the
characteristics of oils produced from microwave pyrolysis of sewage sludge. Fuel
Processing Technology, 86: 1007–1020, 2005.
[158] Miura M., Kaga H., Sakurai A., Kakuchi T., and Takahashi K. Rapid pyrolysis of
wood block by microwave heating. Journal of Analytical and Applied Pyrolysis,
71: 187–199, 2004.
[159] Budarin V.L., Zhao Y., Gronnow M.J., Shuttleworth P.S., Breeden S.W., Macquarrie
D.J., and Clark J.H. Microwave-mediated pyrolysis of macro-algae. Green
Chemistry, 13: 2330, 2011.
[160] Huang Y.F., Kuan W.H., Lo S.L., and Lin C.F. Total recovery of resources
and energy from rice straw using microwave-induced pyrolysis. Bioresource
technology, 99: 8252–8258, 2008.
208 REFERENCES
[161] Yagmur E., Ozmak M., and Aktas Z. A novel method for production of activated
carbon from waste tea by chemical activation with microwave energy. Fuel, 87:
3278–3285, 2008.
[162] Budarin V.L., Clark J.H., Lanigan B.a., Shuttleworth P., Breeden S.W., Wilson
A.J., Macquarrie D.J., Milkowski K., Jones J., Bridgeman T., and Ross A. The
preparation of high-grade bio-oils through the controlled, low temperature
microwave activation of wheat straw. Bioresource technology, 100: 6064–6068,
2009.
[163] El Harfi K., Mokhlisse A., Chanâa M.B., and Outzourhit A. Pyrolysis of the
Moroccan (Tarfaya) oil shales under microwave irradiation. Fuel, 79: 733–742,
2000.
[164] Monsef-Mirzai P. and Ravindran M. The use of microwave heating for the
pyrolysis of coal via inorganic receptors of microwave energy. Fuel, 71: 716–
717, 1992.
[165] Monsef-Mirzai P., Ravindran M., McWhinnie W.R., and Burchill P. Rapid
microwave pyrolysis of coal. Fuel, 74: 20–27, 1995.
[166] Ma S., Zhao Y., Guo T., and Ma X. Experimental Study on Microwave-Induced NO
Decomposition on Activated Carbon Bed. In 2009 Asia-Pacific Power and Energy
Engineering Conference, pp. 1–4. IEEE, 2009.
[167] Fernández Y., Arenillas A., Díez M.A., Pis J.J., and Menéndez J.A. Pyrolysis of
glycerol over activated carbons for syngas production. Journal of Analytical and
Applied Pyrolysis, 84: 145–150, 2009.
[168] Fidalgo B., Fernández Y., Domínguez A., Pis J.J., and Menéndez J. Microwave-
assisted pyrolysis of CH4/N2 mixtures over activated carbon. Journal of
Analytical and Applied Pyrolysis, 82: 162–158, 2008.
[169] Bu Q., Lei H., Ren S., Wang L., Holladay J., Zhang Q., Tang J., and Ruan R. Phenol
and phenolics from lignocellulosic biomass by catalytic microwave pyrolysis.
Bioresource technology, 102: 7004–7, 2011.
[170] Kaminsky W. Thermal recycling of polymers. Journal of Analytical and Applied
Pyrolysis, 8: 439–448, 1985.
[171] Kaminsky W. Pyrolysis of plastic waste and scrap tyres in a fluid bed reactor.
Resource Recovery and Conservation, 5: 205–216, 1980.
[172] Williams E.A. and Williams P.T. The pyrolysis of individual plastics and a plastic
mixture in a fixed bed reactor. Journal of Chemical Technology & Biotechnology,
70: 9–20, 1997.
[173] Williams P.T. and Williams E.A. Fluidised bed pyrolysis of low density
polyethylene to produce petrochemical feedstock. Journal of Analytical and
Applied Pyrolysis, 51: 107–126, 1999.
[174] Conesa J.A., Font R., Marcilla A., and Garcia A.N. Pyrolysis of Polyethylene in a
Fluidized Bed Reactor. Energy & Fuels, 8: 1238–1246, 1994.
[175] Milne B., Behie L., and Berruti F. Recycling of waste plastics by ultrapyrolysis
using an internally circulating fluidized bed reactor. Journal of Analytical and
applied Pyrolysis, 51: 157–166, 1999.
[176] Aguado R., Olazar M., María J., Gaisán B., and Bilbao J. Wax formation in the
pyrolysis of polyolefins in a conical spouted bed reactor. Energy & fuels, 16:
1429–1437, 2002.
[177] Mastral F., Esperanza E., Berrueco C., Juste M., and Ceamanos J. Fluidized bed
thermal degradation products of HDPE in an inert atmosphere and in air-nitrogen
mixtures. Journal of Analytical and Applied Pyrolysis, 70: 1–17, 2003.
REFERENCES 209
[178] Mastral J., Berrueco C., Gea M., and Ceamanos J. Catalytic degradation of high
density polyethylene over nanocrystalline HZSM-5 zeolite. Polymer Degradation
and Stability, 91: 3330–3338, 2006.
[179] Berrueco C. and Mastral F. Production of waxes and tars from the continuous
pyrolysis of high density polyethylene. Influence of operation variables. Energy
& fuels, 16: 1148–1153, 2002.
[180] Ludlow-Palafox C. Enval’s Beginnings. http://www.enval.com/history.php, 2009.
[181] Bockhorn H., Hornung a., Hornung U., and Schawaller D. Kinetic study on the
thermal degradation of polypropylene and polyethylene. Journal of Analytical and
Applied Pyrolysis, 48: 93–109, 1999.
[182] Aguado R., Olazar M., and Gaisán B. Kinetics of polystyrene pyrolysis in a conical
spouted bed reactor. Chemical Engineering . . . , 92: 91–99, 2003.
[183] Scott D.S., Czernik S.R., Piskorz J., and Radlein D.S.A.G. Fast Pyrolysis of Plastic
Wastes. Energy & Fuels, 4: 407–411, 1990.
[184] Bockhorn H., Hentschel J., Hornung a., and Hornung U. Environmental
engineering: Stepwise pyrolysis of plastic waste. Chemical Engineering Science,
54: 3043–3051, 1999.
[185] Kaminsky W. and Franck J. Monomer recovery by pyrolysis of poly (methyl
methacrylate) (PMMA). Journal of Analytical and Applied Pyrolysis, 19: 311–318,
1991.
[186] Kumar S. and Singh R.K. Recovery of Hydrocarbon Liquids from Waste High
Density Polyethylene by Thermal Pyrolysis. Brazilian Journal of Chemical
Engineering, 28: 659–667, 2011.
[187] Campbell I.M. Catalysis at Surfaces. Chapman and Hall, London, 1988.
[188] Sharratt P.N., Lin Y.H., Garforth a.a., and Dwyer J. Investigation of the catalytic
pyrolysis of high-density polyethylene over a HZSM-5 catalyst in a laboratory
fluidized-bed reactor. Industrial & Engineering Chemistry Research, 36: 5118–
5124, 1997.
[189] Miskolczi N. and Bartha L. Investigation of hydrocarbon fractions form waste
plastic recycling by FTIR, GC, EDXRFS and SEC techniques. Journal of
biochemical and biophysical methods, 70: 1247–53, 2008.
[190] Seo Y.H., Lee K.H., and Shin D.H. Investigation of catalytic degradation of high-
density polyethylene by hydrocarbon group type analysis. Journal of Analytical
and Applied Pyrolysis, 70: 383–398, 2003.
[191] Bagri R. and Williams P.T. Catalytic pyrolysis of polyethylene. Journal of
Analytical and Applied Pyrolysis, 63: 29–41, 2002.
[192] Negelein D.L., Lin R., and White R.L. Effects of catalyst acidity and HZSM-5
channel volume on polypropylene cracking. Journal of Applied Polymer Science,
37: 341–348, 1998.
[193] Manos G., Garforth A., and Dwyer J. Catalytic Degradation of High-Density
Polyethylene over Different Zeolitic Structures. Industrial & Engineering
Chemistry Research, 39: 1198–1202, 2000.
[194] Garforth A.A., Lin Y.H., Sharratt P.N., and Dwyer J. Production of hydrocarbons by
catalytic degradation of high density polyethylene in a laboratory fluidised-bed
reactor. Applied Catalysis A: General, 169: 331–342, 1998.
[195] Lin Y.H., Yang M.H., Yeh T.F., and Ger M.D. Catalytic degradation of high
density polyethylene over mesoporous and microporous catalysts in a fluidised-
bed reactor. Polymer Degradation and Stability, 86: 121–128, 2004.
210 REFERENCES
[196] Ali S., Garforth A., Harris D., Rawlence D., and Uemichi Y. Polymer waste
recycling over "used" catalysts. Catalysis Today, 75: 247–255, 2002.
[197] González Y.S., Costa C., Márquez M.C., and Ramos P. Thermal and catalytic
degradation of polyethylene wastes in the presence of silica gel, 5A molecular
sieve and activated carbon. Journal of hazardous materials, 187: 101–112, 2011.
[198] Strouse G., Burns G., Croarkin M., Scroger M., and Guthrie W. Temperature-
Electromotive Force Reference Functions and Tables for the Letter-Designated
Thermocouple Types Based on the ITS-90. Technical report, National Institute of
Standards and Technology, 1993.
[199] King M. Process Control: A Practical Approach. Wiley, 2010.
[200] Ziegler J. and Nichols N. Optimum settings for PID controllers. Transactions of
ASME, 64: 759–768, 1942.
[201] Predel M. and Kaminsky W. Pyrolysis of mixed polyolefins in a fluidised-bed
reactor and on a pyro-GC/MS to yield aliphatic waxes. Polymer Degradation and
Stability, 70: 373–385, 2000.
[202] Kaminsky W. and Zorriqueta I.J. Catalytical and thermal pyrolysis of polyolefins.
Journal of Analytical and Applied Pyrolysis, 79: 368–374, 2007.
[203] Aboulkas A., El harfi K., and El Bouadili A. Thermal degradation behaviors
of polyethylene and polypropylene. Part I: Pyrolysis kinetics and mechanisms.
Energy Conversion and Management, 51: 1363–1369, 2010.
[204] Arabiourrutia M., Elordi G., Lopez G., Borsella E., Bilbao J., and Olazar M.
Characterization of the waxes obtained by the pyrolysis of polyolefin plastics in
a conical spouted bed reactor. Journal of Analytical and Applied Pyrolysis, 94:
230–237, 2012.
[205] Elordi G., Olazar M., Lopez G., Artetxe M., and Bilbao J. Product Yields and
Compositions in the Continuous Pyrolysis of High-Density Polyethylene in a
Conical Spouted Bed Reactor. Industrial & Engineering Chemistry Research,
50: 6650–6659, 2011.
[206] Cozzani V., Nicolella C., Rovatti M., and Tognotti L. Influence of Gas-Phase
Reactions on the Product Yields Obtained in the Pyrolysis of Polyethylene.
Industrial & Engineering Chemistry Research, 36: 342–348, 1997.
[207] Donaj P.J., Kaminsky W., Buzeto F., and Yang W. Pyrolysis of polyolefins for
increasing the yield of monomers’ recovery. Waste management (New York, N.Y.),
32: 840–6, 2012.
[208] Hernández M.d.R., García A.N., and Marcilla A. Catalytic flash pyrolysis of HDPE
in a fluidized bed reactor for recovery of fuel-like hydrocarbons. Journal of
Analytical and Applied Pyrolysis, 78: 272–281, 2007.
[209] McCaffrey W., Cooper D., and Kamal M. Tertiary recycling of polyethylene:
mechanism of liquid production from polyethylene by thermolysis/reactive
distillation. Polymer Degradation and Stability, 62: 513–521, 1998.
[210] Blazsó M. Polyaromatization in common synthetic polymers at elevated
temperatures. Journal of Analytical and Applied Pyrolysis, 25: 25–35, 1993.
[211] Achilias D.S., Roupakias C., Megalokonomos P., Lappas A.A., and Antonakou E.V.
Chemical recycling of plastic wastes made from polyethylene (LDPE and HDPE)
and polypropylene (PP). Journal of hazardous materials, 149: 536–42, 2007.
[212] Al-Salem S. and Lettieri P. Kinetic study of high density polyethylene (HDPE)
pyrolysis. Chemical Engineering Research and Design, 88: 1599–1606, 2010.
REFERENCES 211
[213] Marcilla A., Beltrán M., and Navarro R. Thermal and catalytic pyrolysis of
polyethylene over HZSM5 and HUSY zeolites in a batch reactor under dynamic
conditions. Applied Catalysis B: Environmental, 86: 78–86, 2009.
[214] Hernández M.d.R., Gómez A., García A.N., Agulló J., and Marcilla A. Effect of the
temperature in the nature and extension of the primary and secondary reactions
in the thermal and HZSM-5 catalytic pyrolysis of HDPE. Applied Catalysis A:
General, 317: 183–194, 2007.
[215] Pasel C. and Wanzl W. Experimental investigations on reactor scale-up and
optimisation of product quality in pyrolysis of shredder waste. Fuel Processing
Technology, 80: 47–67, 2003.
[216] Çelikten I., Koca A., and Ali Arslan M. Comparison of performance and emissions
of diesel fuel, rapeseed and soybean oil methyl esters injected at different
pressures. Renewable Energy, 35: 814–820, 2010.
[217] Czernik S. and Bridgwater A.V. Overview of Applications of Biomass Fast
Pyrolysis Oil. Energy & Fuels, 18: 590–598, 2004.
[218] Graboski M. and McCormick R. Combustion of fat and vegetable oil derived fuels
in diesel engines. Progress in Energy and Combustion, 24: 125–164, 1998.
[219] Solantausta Y., Nylund N., and Westerholm M. Wood-pyrolysis oil as fuel in a
diesel-power plant. Bioresource Technology, 46: 177–188, 1993.
[220] ASTM D4814-11b Standard Specification for Automotive Spark-Ignition Engine
Fuel. Technical report, ASTM International, 2011.
[221] ASTM D975-11b Standard Specification for Diesel Fuel Oils. Technical report,
ASTM International, 2011.
[222] Azev V.S., Emel’yanov V.E., and Turovskii F.V. Automotive Gasolines. Long-Term
Requirements for Composition and Properties. Chemistry and Technology of
Fuels and Oils, 40: 291–297, 2004.
[223] Jones D.S. and Pujado P.R. (Eds.). Handbook of Petroleum Processing. Springer
Netherlands, Dordrecht, 2006.
[224] Facts and Figures of the Chemical Industry. Chemical & Engineering News, 90:
59–60, 2011.
[225] Wampler T. Applied Pyrolysis Handbook. CRC Press, Boca Raton, FL, 1st edition,
1995.
[226] Westerhout R.W.J., Waanders J., Kuipers J.A.M., and Swaaij W.P.M.V. Kinetics
of the Low-Temperature Pyrolysis of Polyethene, Polypropene, and Polystyrene
Modeling, Experimental Determination, and Comparison with Literature Models
and Data. Industrial & Engineering Chemistry Research, 36: 1955–1964, 1997.
[227] Kim S. and Kim Y.C. Using isothermal kinetic results to estimate the kinetic triplet
of the pyrolysis of high density polyethylene. Journal of Analytical and Applied
Pyrolysis, 73: 117–121, 2005.
[228] Cypres R. Aromatic hydrocarbons formation during coal pyrolysis. Fuel
Processing Technology, 15: 1–15, 1987.
[229] Rice F.O. The Thermal Decomposition of Organic Compounds from the
Standpoint of Free Radicals. I. Saturated Hydrocarbons. Journal of the American
Chemical Society, 54: 1959–1972, 1931.
[230] Rice F.O., Johnston W., and Evering B. The Thermal Decomposition of Organic
Compounds from the Standpoint of Free Radicals. II. Experimental Evidence of
the Decomposition of Organic Compounds into Free Radicals. Journal of the
American Chemical Society, 54: 3529–3543, 1932.
212 REFERENCES
[231] Rice F.O. The Thermal Decomposition of Organic Compounds from the
Standpoint of Free Radicals. III. The Calculation of the Products Formed from
Paraffin Hydrocarbons. Journal of the American Chemical Society, 55: 3035–
3040, 1933.
[232] Rice F.O. and Dooley M. The Thermal Decompositon of Organic Compounds
from the Standpoint of Free Radicals. IV. The Dehydrogenation of Paraffin
Hydrocarbons and the Strength of the C-C Bond. Journal of the American
Chemical Society, 55: 4245–4247, 1933.
[233] Rice F. and Johnston W. The Thermal Decomposition of Organic Compounds from
the Standpoint of Free Radicals. V. The Strength of Bonds in Organic Molecules.
Journal of the American Chemical Society, 56: 214–219, 1934.
[234] Rice F. and Herzfeld K.F. The Thermal Decomposition of Organic Compounds
from the Standpoint of Free Radicals. VI. The Mechanism of Some Chain
Reactions. Journal of the American Chemical Society, 56: 284–289, 1934.
[235] Kossiakoff A. and Rice F.O. Thermal Decomposition of Hydrocarbons, Resonance
Stabilization and Isomerization of Free Radicals. Journal of the American
Chemical Society, 65: 590–595, 1943.
[236] Simha R. and Wall L.A. Some aspects of depolymerization kinetics. Journal of
Polymer Science, 6: 39–44, 1951.
[237] Wall L.A., Flynn J.H., and Straus S. Rates of Molecular Vaporization of Linear
Alkanes. Journal of Physical Chemistry, 74: 3237–3242, 1970.
[238] Roedel M. The Molecular Structure of Polyethylene. I. Chain Branching in
Polyethylene during Polymerization. Journal of the American Chemical Society,
75: 6110–6112, 1953.
[239] Song C., Lai W.C., and Schobert H.H. Condensed-Phase Pyrolysis of n-
Tetradecane at Elevated Pressures for Long Duration. Product Distribution and
Reaction Mechanisms. Industrial & Engineering Chemistry Research, 33: 534–
547, 1994.
[240] van Schooten J. and Evenhuis J.K. Pyrolysis-Hydrogenation-GLC of Polyolefins.
Polymer, 6: 343–360, 1959.
[241] Wu G., Katsumura Y., Matsuura C., Ishigure K., and Kubo J. Radiation Effect on
the Thermal Cracking of n -Hexadecane. 2. A Kinetic Approach to Chain Reaction.
Industrial & Engineering Chemistry Research, 36: 3498–3504, 1997.
[242] Miller D. Higher Alpha, Omega-Dienes in Paraffin and Olefin Pyrolyzates.
Industrial & Engineering Chemistry Product Research and Development, 2: 220–
223, 1963.
[243] Diels O. and Alder K. Synthesen in der hydroaromatischen Reihe. Justus Liebig’s
Annalen der Chemie, 460: 98–122, 1928.
[244] Crittenden B.D. and Long R. Formation of polycyclic aromatics in rich premixed
acetylene and ethylene flames. Combustion and Flame, 20: 359–368, 1973.
[245] Cole J.A., Bittner J.D., Longwell J.P., and Howard J.B. Formation Mechanisms of
Aromatic Compounds in Aliphatic Flames. Combustion and Flame, 56: 51–70,
1984.
[246] Badger G., Buttery R., Kimber R., Lewis G., Moritz A., and Napier I. The
Formation of Aromatic Hydrocarbons at High Temperatures. Part I. Introduction.
Journal of the Chemical Society, pp. 2449–2452, 1958.
[247] Frenklach M., Clary D.W., Gardiner W.C., and Stein S.E. Detailed kinetic modeling
of soot formation in shock-tube pyrolysis of acetylene. Symposium (International)
on Combustion, 20: 887–901, 1985.
REFERENCES 213
[248] Richter H. and Howard J. Formation of polycyclic aromatic hydrocarbons and
their growth to soot - a review of chemical reaction pathways. Progress in Energy
and Combustion Science, 26: 565–608, 2000.
[249] Cejkaa J., Zilkováa N., Zonesb S., and Bejblováa M. Novel Zeolites in
Transformation of Aromatic Hydrocarbons. 18th Saudi Japan Symposium, 1: 1–9,
2008.
[250] Corma A., Fornes V., Garcia H., Marti V., and Miranda M. Acid Zeolites as
Electron Acceptors. Use of Thianthrene Radical Cation as a Probe. Chemistry
of Materials, 7: 2136–2143, 1995.
[251] Sankararaman S., Yoon K., Yabe T., and Kochi J. Control of back electron transfer
from charge-transfer ion pairs by zeolite supercages. Journal of the American
Chemisty Society, 113: 1419–1421, 1991.
[252] Greensfelder B., Voge H., and Good G. Catalytic and Thermal Cracking of Pure
Hvdrocarbons. Industrial and Engineering Chemistry, 41: 2573–2584, 1949.
[253] Suelves I., Lázaro M., Moliner R., Pinilla J., and Cubero H. Hydrogen production
by methane decarbonization: Carbonaceous catalysts. International Journal of
Hydrogen Energy, 32: 3320–3326, 2007.
[254] Whitmore F.C. Mechanism of the Polymerization of Olefins by Acid Catalysts.
Industrial & Engineering Chemistry, 26: 94–95, 1934.
[255] Corma A. and Orchilles A.V. Current views on the mechanism of catalytic
cracking. Microporous and Mesoporous Materials, 35–36: 21–30, 2000.
[256] Cumming K.A. and Wojciechowski B.W. Hydrogen Transfer, Coke Formation, and
Catalyst Decay and Their Role in the Chain Mechanism of Catalytic Cracking.
Catalysis Reviews: Science and Engineering, 38: 101–157, 2006.
[257] Zawadzki J. IR spectroscopy studies of oxygen surface compounds on carbon.
Carbon, 16: 491–497, 1978.
[258] Ishizaki C. and Marti I. Surface oxide structures on a commercial activated
carbon. Carbon, 19: 409–412, 1981.
[259] Strelko Jr. V., Malik D.J., and Streat M. Characterisation of the surface of oxidised
carbon adsorbents. Carbon, 40: 95–104, 2002.
[260] Kolthoff I.M. Properties of Active Charcoal Reactivated in Oxygen at 400A.
Journal of the American Chemical Society, 54: 4473–4480, 1932.
[261] King A. Studies in chemisorption on charcoal. Part IX. The influence of
temperature of activation on the sorption of acids and bases. Journal of the
Chemical Society (Resumed), pp. 1489–1491, 1937.
[262] Brunauer S., Emmett P.H., and Teller E. Adsorption of Gases in Multimolecular
Layers. Journal of the American Chemical Society, 60: 309–319, 1938.
[263] Barrett E., Joyner L., and Halenda P. The Determination of Pore Volume and Area
Distributions in Porous Substances. I. Computations from Nitrogen Isotherms.
Journal of the American Chemical Society, 73: 373–380, 1951.
[264] Sing K.S.W., Everett D.H., Haul R.A.W., Moscou L., Pierotti R.A., Rouquerol J., and
Siemieniewska T. Reporting physisorption data for gas/solid systems with special
reference to the determination of surface area and porosity (Recommendations
1984). Pure and Applied Chemistry, 57: 603–619, 1985.
[265] Rouquerou J., Avnir D., Fairbridge C.W., Everett D.H., Haynes J.H., Pernicone
N., Ramsay J.D.F., Sing K.S.W., and Unger K.K. Recommendations for the
characterization of porous solids. Pure and Applied Chemistry, 66: 1739–1758,
1994.
214 REFERENCES
[266] Rouquerol F., Rouquerol J., and Sing K. Adsorption by Powders and Porous Solids.
Principles, Methodology and Applications. null. Academic Press, San Diego,
1999.
[267] Marsh H. Adsorption methods to study microporosity in coals and carbons - a
critique. Carbon, 25: 49–58, 1987.
[268] Menéndez J.A., Menéndez E.M., Garcia A., Parra J.B., and Pis J.J. Thermal
treatment of active carbons: A comparison between microwave and electrical
heating. Journal of Microwave Power and Electromagetic Energy, 34: 137–143,
1999.
[269] Marsh H., Crawford D., O’Grady T., and Wennerberg A. Carbons of high surface
area. A study by adsorption and high resolution electron microscopy. Carbon,
20: 419–426, 1982.
[270] Gamby J., Taberna P., Simon P., Fauvarque J., and Chesneau M. Studies
and characterisations of various activated carbons used for carbon/carbon
supercapacitors. Journal of Power Sources, 101: 109–116, 2001.
[271] Otowa T., Nojima Y., and Miyazaki T. Development of KOH activated high surface
area carbon and its application to drinking water purification. Carbon, 35: 1315–
1319, 1997.
[272] Evans M.J.B., Halliop E., and Macdonald J.A.F. The production of chemically-
activated carbon. Carbon, 37: 269–274, 1999.
[273] Kaneko K. Determination of pore size and pore size distribution 1. Adsorbents
and catalysts. Journal of Membrane Science, 96: 59–89, 1994.
[274] Ismadji S. and Bhatia S.K. Effect of pore-network connectivity on multicomponent
adsorption of large molecules. American Institute of Chemical Engineers Journal,
49: 65–81, 2003.
[275] Dubinin M. Investigation on the nature of the microporous structure in activated
charcoals. Russian Chemical Bulletin, 10: 693–698, 1961.
[276] Skalny J., Bodor E.E., and Brunauer S. Investigations of a Complete Pore-
Structure Analysis III . Analysis of Carbon Adsorbents. Journal of Colloid and
Interface Science, 37: 478–483, 1971.
[277] Moreira R., José H., and Rodrigues A. Modification of pore size in activated
carbon by polymer deposition and its effects on molecular sieve selectivity.
Carbon, 39: 2269–2276, 2001.
[278] Margulès L., Demaison J., and Boggs J. The Equilibrium CC Bond Length.
Structural Chemistry, 11: 145–154, 2000.
[279] Freitas M.M.A. and Figueiredo J.L. Preparation of carbon molecular sieves for
gas separations by modification of the pore sizes of activated carbons. Fuel, 80:
1–6, 2001.
[280] Foley H.C. Carbogenic molecular sieves: synthesis, properties and applications.
Microporous Materials, 4: 407–433, 1995.
[281] Gonçalves M., Molina-Sabio M., and Rodriguez-Reinoso F. Modification of
activated carbon hydrophobicity by pyrolysis of propene. Journal of Analytical
and Applied Pyrolysis, 89: 17–21, 2010.
[282] Rouzaud J.N. and Clinard C. Quantitative high-resolution transmission electron
microscopy: a promising tool for carbon materials characterization. Fuel
Processing Technology, 77-78: 229–235, 2002.
REFERENCES 215
[283] Font R., Aracil I., Fullana A., Martin-Gullon I., and Conesa J. Semivolatile
compounds in pyrolysis of polyethylene. Journal of Analytical and Applied
Pyrolysis, 68-69: 599–611, 2003.
[284] Forzatti P. and Lietti L. Catalyst Deactivation. Catalysis Today, 52: 165–181,
1999.
[285] Bibby D.M., Howe R.F., and Mclellan G.D. Coke formation in high-silica zeolites.
Applied Catalysis A: General, 93: 1–34, 1992.
[286] Froment G.F. and Bischoff K.B. Non-steady state behaviour of fixed bed catalytic
catalyst fouling. Chemical Engineering Science, 16: 189–201, 1961.
[287] Fidalgo B., Fernández Y., Zubizarreta L., Arenillas A., Domínguez A., Pis J.,
and Menéndez J. Growth of nanofilaments on carbon-based materials from
microwave-assisted decomposition of CH4. Applied Surface Science, 254: 3553–
3557, 2008.
[288] Serrano D., Botas J., and Guil-Lopez R. H2 production from methane pyrolysis
over commercial carbon catalysts: Kinetic and deactivation study. International
Journal of Hydrogen Energy, 34: 4488–4494, 2009.
[289] Johnson C.A. and Thomas K. Applications of Raman microprobe spectroscopy to
the characterization carbon deposits on catalysts. Fuel, 63: 1073–1080, 1984.
[290] Prasetyo I. and Do D. Pore structure alteration of porous carbon by catalytic coke
deposition. Carbon, 37: 1909–1918, 1999.
[291] Ania C., Menéndez J., Parra J., and Pis J. Microwave-induced regeneration of
activated carbons polluted with phenol. A comparison with conventional thermal
regeneration. Carbon, 42: 1383–1387, 2004.
[292] Ania C., Parra J., Pevida C., Arenillas a., Rubiera F., and Pis J. Pyrolysis of
activated carbons exhausted with organic compounds. Journal of Analytical and
Applied Pyrolysis, 74: 518–524, 2005.
[293] Ania C., Parra J., Menéndez J., and Pis J. Effect of microwave and conventional
regeneration on the microporous and mesoporous network and on the adsorptive
capacity of activated carbons. Microporous and Mesoporous Materials, 85: 7–15,
2005.
[294] Bo L.L., Wang X.C., and Lu L. Study on the purification and regeneration of
activated carbon by microwave and its adsorption capacity for Acid Orange 7.
Xi’an Jianzhu Keji Daxue Xuebao/Journal of Xi’an University of Architecture and
Technology, 40: 413–417, 2008.
[295] Chang S.H., Wang K.S., Liang H.H., Chen H.Y., Li H.C., Peng T.H., Su Y.C.,
and Chang C.Y. Treatment of Reactive Black 5 by combined electrocoagulation-
granular activated carbon adsorption-microwave regeneration process. Journal
of hazardous materials, 175: 850–7, 2010.
[296] Wang J., Peng X., Luan Z., and Zhao C. Regeneration of carbon nanotubes
exhausted with dye reactive red 3BS using microwave irradiation. Journal of
hazardous materials, 178: 1125–1127, 2010.
[297] Cha C.Y., Wallace S., George A.H., and Rogers S. Microwave Technology for
Treatment of Fume Hood Exhaust. Journal of Environmental Engineering, 130:
338–348, 2004.
[298] Hashisho Z., Rood M., and Botich L. Microwave-swing adsorption to capture and
recover vapors from air streams with activated carbon fiber cloth. Environmental
science & technology, 39: 6851–6859, 2005.
216 REFERENCES
[299] Kim K.J. and Ahn H.G. A study on adsorption characteristics of benzene over
activated carbons coated with insulating materials and desorption by microwave
irradiation. Journal of the Korean Industrial and Engineering Chemistry, 19:
445–451, 2008.
[300] Norman L.M. and Cha C.Y. Production of Activated Carbon from Coal Chars using
Microwave Energy. Chemical Engineering Communications, 140: 87–110, 1995.
[301] Guo J. and Lua A.C. Preparation of activated carbons from oil-palm-stone chars
by microwave-induced carbon dioxide activation. Carbon, 38: 1985–1993, 2000.
[302] Valente Nabais J., Carrott P., Ribeiro Carrott M., and Menéndez J. Preparation
and modification of activated carbon fibres by microwave heating. Carbon, 42:
1315–1320, 2004.
[303] Lee K.H., Jeon S.G., Kim K.H., Noh N.S., Shin D.H., Park J., Seo Y., Yee J.J., and
Kim G.T. Thermal and catalytic degradation of waste high-density polyethylene
(HDPE) using spent FCC catalyst. Korean Journal of Chemical Engineering, 20:
693–697, 2003.
[304] Padleckas H. Tray Distillation Tower. http://commons.wikimedia.org/wiki/
File:Tray_Distillation_Tower_EN.svg, 2010.
[305] Artetxe M., Lopez G., Amutio M., Elordi G., Bilbao J., and Olazar M. Light olefins
from HDPE cracking in a two-step thermal and catalytic process. Chemical
Engineering Journal, In Press, 2012.