Sie sind auf Seite 1von 208

Feedstock Recycling of Plastic Wastes

RSC Clean Technology Monographs


Series Editor: J.H. Clark, University of York, U K
Advisory Panel: R.S. Drago (Gainesville, USA), N.M. Edinberry (Sandwich, UK), J. Emsley (London, UK), S.M. Hassur (Washington DC, USA), D.R. Kelly (Cardifi UK), T. Laird (Mayfield, UK), T. Papenfuhs (Frankfurt, Germany), B. Pearson (Wigan, UK),J. Winfield (Glasgow, UK) The chemical process industries are under increasing pressure to develop environmentally friendly products and processes, with the key being a reduction in waste. This timely new series will introduce different clean technology concepts to academics and industrialists, presenting current research and addressing problem-solving issues. Feedstock Recycling of Plastic Wastes by J. Aguado, Rey Juan Carlos University,Mdstoles, Spain; D.P. Serrano, Complutense University of Madrid, Spain Applications of Hydrogen Peroxide and Derivatives by C.W. Jones, formerly of Solvay Interox R & D , Widnes, U K

How to obtain future titles on publication A standing order plan is available for this series. A standing order will bring delivery of each new volume upon publication. For further information please contact: Sales and Customer Care Royal Society of Chemistry Thomas Graham House Science Park Milton Road Cambridge CB4 OWF Telephone: + 44(0) 1223 420066

CLEAN TECHNOLOGY
MONOGRAPHS

Feedstock Recycling of Plastic Wastes

Josk Aguado Department of Experimental Sciences and Engineering, Rey Juan Carlos University, Mdstoles, Spain

David P. Serrano Chemical Engineering Department, Complutense University of Madrid, Spain

RSaC
ROYAL SOCIETY OF CHEMISTRY

ISBN 0-85404-531-7
A catalogue record for this book is available from the British Library

0The Royal Society of Chemistry 1999


All rights reserved. Apart from any fair dealing for the purposes o f research or private study, or criticism or review as permitted under the terms of the UK Copyright, Designs and Patents Act, 1988, this publication may not be reproduced, stored or transmitted, in any form or by any means, without the prior permission in writing of The Royal Society of Chemistry, in the case of reprographic reproduction only in accordance with the terms of the licences issued by the f the licences Copyright Licensing Agency in the UK, or in accordance with the terms o issued by the appropriate Reproduction Rights Organization outside the UK. Enquiries concerning reproduction outside the terms stated here should be sent to The Royal Society of Chemistry at the address printed on this page.

Published by The Royal Society of Chemistry, Thomas Graham House, Science Park, Milton Road Cambridge CB4 OWF, UK For further information see our web site at www.rsc.org Typeset by Paston PrePress Ltd, Beccles, Suffolk Printed by MPG Books Ltd, Bodmin, Cornwall

Preface
The use of plastic materials in daily life has continuously increased over the last 30 years. The amount of plastic consumed per inhabitant in the industrialized countries has increased by a factor of 60 over this period, while the generation of plastic wastes has grown at a similar rate. Thus, over 17.5 million tonnes of plastic wastes are generated per year in Western Europe, their environmental impact being a matter of great public concern. The variation in properties and chemical composition between different types of plastic materials hinders the application of an integrated and general approach to handling these plastic wastes. The light weight of plastic goods, and the fact that plastic wastes are mainly found in MSW (municipal solid waste) mixed with other classes of residues, are factors that greatly limit their recycling. As a consequence, the primary destination of plastic wastes is landfill sites, where they remain for decades due to their slow degradation. In 1996, only around 10% of the plastic wastes generated in Europe were recycled, whereas over 70% were disposed of in landfills. At present, there are three main alternatives for the management of plastic wastes in addition to landfilling: (i) mechanical recycling by melting and regranulation of the used plastics, (ii) feedstock recycling and (iii) energy recovery. Mechanical recycling is limited both by the low purity of the polymeric wastes and the limited market for the recycled products. Recycled polymers only have commercial applications when the plastic wastes have been subjected to a previous separation by resin; recycled mixed plastics can only be used in undemanding applications. On the other hand, energy recovery by incineration, although an efficient alternative for the removal of solid wastes, is the subject of great public concern due to the contribution of combustion gases to atmospheric pollution. There has also been some controversy in the past about the possible relationship between dioxin formation and the presence of C1-containing plastics in the waste stream. Consequently, feedstock recycling appears as a potentially interesting approach, based on the conversion of plastic wastes into valuable chemicals useful as fuels or as raw materials for the chemical industry. The cleavage and degradation of the polymer chains may be promoted by temperature, chemical agents, catalysts, etc. The aim of this work is to describe and review the different alternatives developed for the feedstock recycling of plastic wastes, with emphasis on both the scientific and technical aspects. Due to the wide variety of plastic types, the
V

vi

Preface

work focuses on the major polymers present in household and industrial plastic wastes: polyethylene (PE), polypropylene (PP), polystyrene (PS), polyethylene terephthalate (PET), polyvinyl chloride (PVC), polyurethanes (PU) and polyamides (PA). These plastics account for more than 90% of total plastic wastes. Although elastomers are not usually considered as plastic materials, the book also covers the feedstock recycling of rubber wastes, mainly used tyres. This is supported by the fact that a number of degradation treatments have been developed which can be used for both plastic and rubber wastes. Five main types of feedstock recycling processes have been considered: chemical depolymerization, gasification and partial oxidation, thermal degradation, catalytic cracking and reforming, and hydrogenation. Each of these alternatives is reviewed in an independent chapter, highlighting the most recent progress with extensive literature references. Besides conventional treatments (pyrolysis, gasification, etc.), the book includes new technological approaches for the degradation of plastics such as conversion under supercritical conditions and coprocessing with coal. The first chapter gives a general introduction to the types and applications of polymeric materials, as well as to the various plastic waste management and recycling alternatives. Data are provided about the volume of plastic wastes generated, their origin and their composition. Previous separation and classification of the plastics is required in many feedstock recycling processes, and so the different methods available for plastic sorting are described: manual, density differences, selective dissolution, automated methods based on spectroscopic techniques, etc. Chapter 2 discusses depolymerization processes based on the chemical cleavage of polymer molecules to convert them back into the raw monomers. The latter can be reused in the manufacture of new polymers, with properties similar to those of the virgin resins. However, this alternative is mainly used for condensation polymers, and is not successful for the degradation of most addition polymers. Glycolysis, methanolysis, hydrolysis and ammonolysis are the main treatments considered. Chemical depolymerization of polyesters, polyurethanes and pol yamides is reviewed. Chapter 3 deals mainly with gasification processes leading to synthesis gas, which is a mixture useful for the preparation of a variety of chemical products (ammonia, methanol, hydrocarbons, etc.). Gasification processes based on treatment with oxygen, air and steam are described. In many cases, gasification of plastic wastes takes place simultaneously with that of other organic residues, coal and petroleum fractions. In addition to gasification, other degradation alternatives based on partial oxidation methods are described in this chapter. The degradation of plastic wastes by thermal treatments in the absence of oxygen is reviewed in Chapter 4. Depending on the raw polymer and the degradation conditions, a variety of thermal processes have been considered: thermal depolymerization into the raw monomers, thermal cracking, pyrolysis, steam cracking and thermal treatment in the presence of solvents. For each treatment both the products derived and the different types of reactors used are described.

Preface

vii

Chapter 5 is devoted to catalytic processes for plastic waste recycling. Through selection of the right catalysts, the plastic degradation can be used to obtain a number of valuable products. The properties of the main types of catalysts are reviewed. Both direct catalytic cracking processes, and the combination of a previous thermal cracking of the plastic wastes with a catalytic reforming of the gases generated in the former are considered. Chapter 6 deals with hydrogenation processes, usually based on the use of bifunctional catalysts. Plastic and rubber degradation in a hydrogen atmosphere is an effective treatment yielding highly saturated oils. Coliquefaction of plastics or rubber with coal is also considered. The last chapter highlights the main conclusions and establishes a comparative study of the various alternatives for the feedstock recycling of plastic wastes. The final conclusion is that feedstock recycling of both plastic and rubber wastes has a high potential for growth in the next few years, although to be commercially successful a number of technical and economic aspects still have to be addressed.

Acknowledgements
We are indebted to those who took the time to help us in drawing many of the figures and schemes in this book: Raul Sanz, Josk M. Escola, Rafael Garcia, Luis M. Garcia and Araceli Rodriguez. We also greatly appreciate the efforts and work of Prof. Rafael van Grieken, who patiently reviewed all the chapters. Finally, we would like to thank all those colleagues that gave permission for the reproduction of figures from their work.

Contents
Preface Acknowledgements
Vlll

...

Chapter 1 Introduction
1 Significance of Plastic Materials in Todays Society 2 Classes of Organic Polymers and their Main Applications Classification of Polymers

1
1 3 4 8 11 11 12 13

Thermoplastics Thermosets Plastic Additives Rubber Plastic Wastes The Economic and Environmental Impact of Plastic Wastes Management of Plastic Wastes Mechanical Recycling Feedstock Recycling Sorting and Separation of Mixed Plastics Future Trends in Plastic Waste Management References

15 16
19 20 22 27 28

Chapter 2 Chemical Depolymerization


1 Introduction 2 Polyesters

31
31 32 33 37 38 41 42 44 45 46 47 49 50

Glycolysis Methanolysis Hydrolysis Ammonolysis and Aminolysis Combined Chemolysis Methods Comparison of the Various PET Chemolysis Methods Polyurethanes Glycolysis Hydrolysis Ammonolysis and Aminolysis Combined Chemolysis Methods
ix

Contents

4 5 6

Polyamides, Polycarbonates and Polyacetals Summary References

52 55 56

Chapter 3 Gasification and Partial Oxidation


1 Introduction 2 Gasification of Carbonaceous Materials and Uses of the Syngas 3 Gasification of Plastic and Rubber Wastes 4 Gasification of Mixed Solid Wastes 5 Other Plastic and Rubber Partial Oxidation Processes 6 Summary 7 References

59
59 59 62 67 69 71 71

Chapter 4 Thermal Processes

73

73 1 Introduction 2 Mechanism of the Thermal Degradation of Addition 74 Polymers 77 3 Thermal Conversion of Individual Plastics 78 Polyethylene 85 Polypropylene 86 Polystyrene 91 Polyvinyl Chloride 98 Other Plastics 100 4 Thermal Conversion of Plastic Mixtures Interactions Between Components During Thermal 101 Degradation Processes for the Thermal Degradation of Plastic Wastes 105 5 Thermal Coprocessing of Plastic Wastes with Coal and 115 Lignocellulosic Materials 117 6 Thermal Conversion of Rubber Wastes and Used Tyres 122 7 Summary 124 8 References

Chapter 5 Catalytic Cracking and Reforming


1 Introduction 2 Types and Properties of Polymer Cracking Catalysts 3 Catalytic Conversion of Individual Plastics Polyethylene Polypropylene Polystyrene 4 Catalytic Conversion of Plastic Mixtures and Rubber Wastes

129
129 130 133 133
145

148 150

Con tents

xi

5
6 7

Conversion of Plastics by a Combination of Thermal and Catalytic Treatments Summary References

151 157 158

Chapter 6 Hydrogenation
1 2 3 4 5 6 7 Introduction Hydrocracking of Plastics Hydrocracking of Rubber and Used Tyres Coliquefaction of Coal and Plastics Coliquefaction of Coal with Rubber and Used Tyres Summary References

161
161 161 168 171 173 176 177

Chapter 7 Concluding Remarks Subject Index

179
185

To Maribel and to Juany

CHAPTER 1

Introduction
1 Significance of Plastic Materials in Todays Society
Plastics are not, as many people believe, new materials. Their origin can be traced to 1847 when Shonbein produced the first thermoplastic resin, celluloid, by reaction of cellulose with nitric acid. However, the general acceptance and commercialization of plastics began during the Second World War when natural polymers, such as natural rubber, were in short supply. Thus, polystyrene was developed in 1937, low density polyethylene in 1941, whereas other commodity plastics such as high density polyethylene and polypropylene were introduced in 1957. Today, plastics are very important materials having widespread use in the manufacture of a variety of products including packaging, textiles, floor coverings, pipes, foams, and car and furniture components. Plastics are synthesized mainly from petroleum-derived chemicals, although only about 4% of total petroleum production is used in the manufacture of plastics. The main reasons for the continuous increase in the demand for commodity plastics are as follows:
0

Plastics are low density solids, which makes it possible to produce lightweight objects. Plastics have low thermal and electric conductivities, hence they are
widely used for insulation purposes.

Plastics are easily moulded into desired shapes. Plastics usually exhibit high corrosion resistance and low degradation rates and are highly durable materials. Plastics are low-cost materials.

Engineering plastics, particularly thermosets, are also used in composite materials. Their excellent technological properties make them suitable for applications in cars, ships, aircraft, telecommunications equipment, etc. In recent years, important new areas of application for plastics have emerged in medicine (fabrication of artificial organs, orthopaedic implants, and devices for the controlled release of drugs), electronics (development of conductive poly1

Chapter I

mers for semiconductor circuits, conductive paints, and electronic shielding), and computer technology (use of polymers with non-linear optical properties for optical data storage). The above paragraphs show that today plastic materials are used in almost all areas of daily life. Accordingly, the production and transformation of plastics are major worldwide industries. Consumption of plastics in Western Europe is forecast to grow from 24.9 million tonnes in 1995 up to about 37 million tonnes in 2006,' an annual growth rate of 4%. This prediction places plastics among the most important materials in the next century also. Table I . I summarizes the changes in total plastic consumption in Western Europe from 1992 to 1996.* These data refer to the final market for plastic products consumed by end-users but they do not include sectors such as textile fibres, elastomers, coatings, or products in which plastics are present in small quantities, because these are not considered as plastic products. If non-plastic applications are also taken into account, the total plastic consumption in Western Europe in 1996 increases up to 33.4 million tonnes. By comparison, the consumption of plastics in the USA and Japan in 1995 were 33.9 and 11.3 million tonnes, respective~y.~ The main sectors of plastic consumption in Western Europe are shown in Figure 1.1. The major field of plastic consumption is packaging, accounting for more than 40% of the total volume, followed by the building and automotive sectors. The most important uses of plastics in packaging are the production of films and sheets, sacks, bags, bottles and foams. In the building sector, plastics are used in a variety of applications: insulation, floor and wall coverings, window and door profiles, pipes, etc. The automotive sector is a good example of the continuous increase in the use of plastic materials. A car's weight can be reduced by 100-200 kg through the replacement of conventional metallic materials by plastics. Fuel tanks, bumpers, bonnets, insulation, seats, dashboards, textiles, batteries, etc. are examples of car components commonly manufactured with plastic materials. Plastics are used for a variety of applications in the agricultural sector such as greenhouses, tunnel and silage films,
pipes for both drainage and irrigation, drums and tanks, etc.

Figure 1.2 illustrates plastic consumption in Western Europe by product for 1995,4confirming that plastics are versatile materials which can be found in a wide range of products. The production and consumption of plastics have continuously increased over recent decades. The plastic consumption per capita in Western Europe has increased from 1 kg per inhabitant in 1960 to about 65 kg per inhabitant in 1995.

Table 1.1 Total plastic consumption in Western Europe (based on reference 2)


Year
Plastic consumption (million tonnes) 1992 24 730 1993 24 360 1994 26 260 1995 24 909 1996 25 905

In t r oduct ion
Others 30% Automoth 7%

Building 19% Agriculture 2%

Figure 1.1 Plastic consumption by sector in Western Europe (1996) .2

TOTAL: 24.9 Mtonnes


25.00%

Figure 1.2 Plastic consumption by product in Western Europe (1995).4

2 Classes of Organic Polymers and their Main


Applications
Polymers are long-chain molecules composed of a large number of identical units called repeating units. A polymer can be expressed as follows:

where RU is the repeating unit and n the number of units present in the polymer molecule. The number of repeating units must be large enough that no variations in the polymer macroscopic properties occur by small changes in the

Chapter I

number of repeating units. This concept enables a distinction to be made between polymers and oligomers. Oligomers are molecules with a small number of repeating units, hence their properties vary significantly by just adding or removing a repeating unit. Most of the polymers with commercial applications are synthetic materials. They are prepared by polymerization reactions involving the chemical linkage of small individual molecules (monomers) to give long-chain polymeric molecules. In some cases, polymers are synthesized by reaction between several monomers. The product so obtained is called a copolymer while the starting molecules are known as comonomers. The structure of copolymers depends on both the relative proportion and the sequence of the different comonomers along the macromolecular chain. Depending on the polymerization conditions, it is possible to obtain random, alternating, block or graft copolymers, as illustrated in Figure 1.3. It is not easy to define the term plastic, which is usually considered as equivalent to the term polymer. Plastics are polymeric materials, but not all polymers are plastics. In general, the term plastic is used to refer to any commercial polymeric material other than fibres and elastomers. Moreover, commercial plastics include other components such as additives, fillers, and a variety of compounds incorporated into the polymers to improve their properties. The term resin is usually used to describe the virgin polymeric material without any of these components.

Classification of Polymers
Polymers are commonly classified according to two main criteria: thermal behaviour and polymerization mechanism. As explained further below, these classifications are important from the point of view of polymer recycling, because the most suitable method for the degradation of a given polymer is closely related to both its thermal properties and its polymerization mechanism.
8)

-A-B-B-A-A-A-B-A-B-B-8-A-B-A-A-B-A-A-B-

b)
C)

-A-B-A-B-A-B-A-B-A-B-A-B-A-B-A-B-A-A-A-A-A-A-0-0-0-0-B-0-B-A-A-A-A-A-

?
d)

-A-A-A-A-A-A-A-A-A-A-A-A-A-A-

Figure 1.3

Possible structures o f copolymers containing A and B repeating units: (a) random, (b) alternating, (c)block, (d) graft.

Introduction

Class@ntion According to Thermal Behaviour


Plastics are divided into two major groups depending on their behaviour when they are heated:
0

Thermoplastics are plastics which undergo a softening when heated to a particular temperature. This thermoplastic behaviour is a consequence of the absence of covalent bonds between the polymeric chains, which remain as practically independent units linked only by weak electrostatic forces (Figure 1.4(a)). Therefore, waste thermoplastics can be easily reprocessed by heating and forming into a new shape. From a commercial point of view, the most important thermoplastics are high density polyethylene (HDPE), low density polyethylene (LDPE), polypropylene (PP), polystyrene (PS), polyvinyl chloride (PVC), polyethylene terephthalate (PET), polyamide (PA), polymethyl methacrylate (PMMA), acrylonitrile-butadiene-styrene copolymer (ABS), and styrene-acrylonitrile copolymer (SAN). Thermosets are plastics whose polymeric chains are chemically linked by strong covalent bonds, which lead to three-dimensional network structures (see Figure 1.4(b)). Once formed into a given shape, thermosets

a) Thermoplastic

b) Thermoset

Figure 1.4 Schematic structures o f thermoplastic and thermoset polymers.

Chapter 1

cannot be reprocessed or remoulded by heating. Examples of thermosets with significant commercial applications are polyurethanes (PU), epoxy resins, unsaturated polyesters and phenol-formaldehyde resins. Thermosets are produced in smaller amounts than thermoplastics, as can be seen in Table 1.2. Table 1.2 summarizes the production of different plastics in Western Europe over the period 1994-1996. Thermosets account for just 16% of total plastic production. A similar ratio of thermoset to thermoplastic production is found in the USA. Elastomers constitute a third class of polymers. Similarly to thermosets, elastomers have a network structure formed by crosslinking between the polymer chains. However, the number of links is less than in the case of thermosets which gives these materials elastic properties. Elastomers can be deformed by the application of external forces. When these forces are suppressed, the polymer recovers its original form. From a commercial point

Table1.2 Consumption of plastics in Western Europe by resin (based on reference 2 )


Consumption ( k tonnes per year)

Resin
LDPE/LLDPE HDPE PP PVC PS/EPS PET ABS/SAN PMMA Acetals Polycar bonates PA Acrylics Others Total thermoplastics UF/MF resins PU Phenolic resins Unsaturated polyesters Alkyd resins Epoxy resins Total thermosets
Total plastics

1994
5825 3718 4982 540 1 2352 1971 550 234 104 202 993 433 29 1 27 056 1174 1720 600 436 340 257 4527
31 583

1995
5723 3613 5020 5300 2525 2101 567
270

I996
5969 3846 5397 5406 255 1 2093 540
257

110 225 1028 34 1 232 27 055 1915 1755 610 430 343 270 5323
32 378

111 233 1027 349 232 28011 1934 1773 610 420 343 275 5355
33 366

Introduction
Thermodrstics Thermosets

PE PP
ww

Polyurethanes
w#O-R-O-CO-NH-R'-NH-COM

Pol yureas
NH -CO -NH-R
> , w b

PS

* CH2-CH
I

rn

Epoxy resins

PVC

Phenolic resins

PET

Unsatured Polyesters PMMA


wCO-CH=CH-CO-O-R -0
*VA

Elastomers

PA

w NH-R-NH-CO-R'-CO AW

Natural Rubber

Polycarbonates Pol yacetals

0-CO -0- R rn

Polybutadiene
wCH2-0-CH2-0
'MN

CH*-CH=CH-CHz

'*

Figure 1.5 Repeating units o f diferent polymers.

of view, rubbers are the main class of elastomers, being mainly used in the manufacture of tyres. The repeating units corresponding to a variety of organic polymers are shown in Figure 1.5.

ClasslJicat ion A ccording to Polymerization Mechanism


Depending on the mechanism of polymerization, two groups of plastic materials can be identified:
Addition polymers. The polymerization proceeds by a sequential incorporation of monomeric molecules into the growing polymer chain, without the release of any molecules or fragments to the reaction medium. As a consequence, the repeating units of addition polymers have the same

Chapter I

chemical composition as the monomers. Examples of addition polymers include PE, PS, PVC, PMMA, etc. Condensation polymers. In this case the polymerization reactions take place with the liberation of small molecules, such as water, hydrochloric acid, etc. Nylon-6,6, obtained by polycondensation of adipic acid and hexamethylenediamine is a classic example of a condensation polymer. As shown in Figure 1.6, this polymerization reaction proceeds with the release of two water molecules by each repeating unit.

Thermoplastics
Thermoplastics account for the majority of plastics consumption. They are used in a wide variety of products and applications. It can be seen from Table 1.2 that about 90% of the total thermoplastics consumption in Western Europe corresponds to just five thermoplastics: PE, PP, PVC, PS and PET. The main properties of these resins are briefly described below.

Polyethylene ( P E )
Polyethylene is synthesized by polyaddition of ethylene molecules, which leads to different types of PE depending on the reaction conditions:
0

High density polyethylene (HDPE) is produced at relatively low temperature (60-200 "C) and pressure (1-100 atm) and is a highly linear polymer having a specific gravity in the range 0.94-0.97 and a high degree of crystallinity (8095%). The main applications of HDPE are for the manufacture of films, food and domestic containers, crates, toys, gas tanks, pipes, etc. by blow moulding and injection moulding. The production of blown films for bags accounts for about 7 % of the HDPE market.

n HO-CfCH21f4C-OH
Adipic acid

n H2NfCH2fNH2 6
Hexame thylenediam ine

1
Nylon-6,6
( repeating unit )

+ 2nH20

Figure 1.6 Synthesis of nylon-6,6 by condensation polymerization.

Introduction

Ultrahigh molecular weight polyethylene (UHMWPE) is really a variety of HDPE with a molecular weight greater than 3 x lo6. UHMWPE is a strong and lightweight plastic used in the fibre industry and for specialized applications such as its use in medicine for the manufacture of artificial hips. Low density polyethylene (LDPE). Unlike HDPE, this type of polyethylene is synthesized at very high pressures (1200-1500 atm) and at temperatures of about 250C. LDPE is a highly branched polymer characterized by its lower crystallinity and specific gravity than HDPE but with greater flexibility. Both the flexibility and crystallinity of LDPE can be controlled by adding low concentrations of acryl or vinyl monomers during the polymerization. LDPE has widespread use in films for bags and food packaging, greenhouses, bottles, cable insulation and injection moulded products. Linear low density polyethylene (LLDPE) is synthesized by copolymerization of ethylene and a-olefins, mainly 1-butene and 1-hexene. The role of the a-olefinic comonomers is to control both the number and the length of the side branches. As a consequence, LLDPE is a polymer with intermediate properties with respect to LDPE and HDPE. Main applications for LLDPE are films, injection moulded parts and wire insulation.

Polypropylene ( P P )
Polypropylene is synthesized by polymerization of propylene, which may result in two main types of PP with commercial applications:
0

Isotactic polypropylene (1-PP) is the most widely produced type. In this polymer, all the pendant methyl groups are located on the same side of the backbone, which results in a high crystallinity ( 8 0 4 5 % ) . Isotactic polypropylene is synthesized at temperatures in the range 50-80 "C and at pressures of 5-25 atm. The main commercial applications of 1-PP are the manufacture of injection moulded containers, pipes, sheets and textile fibres for carpets. 1-PP is more rigid and crack resistant than HDPE, having good electrical insulation properties. Moreover, i-PP has a higher crystalline melting temperature (Tm)which enables its use in products that must be steam sterilized. These facts explain the continuous increase in the use of i-PP in various sectors. Syndiotactic polypropylene (s-PP) is produced at lower temperatures than 1-PP in the presence of Ziegler-Natta catalysts. The side methyl groups in this case are in alternating positions along the chain, which results in a non-crystalline polymer with lower density, mechanical strength and Tm than 1-PP. Accordingly, s-PP is consumed in significantly lower amounts, being used as a coating material and in hot melt adhesives.

10

Chapter I

Polystyrene ( P S )
Polystyrene is produced by styrene monomer polymerization, which leads to an amorphous, non-flexible polymer having good electrical insulation properties and a density of about 1.04 g/cm3. However, its high brittleness and low softening temperature ( < 100 "C) are important limitations on its industrial application. PS is used in the manufacture of radio and TV parts, toys, electronic components, etc. Expanded polystyrene (EPS) is prepared by impregnation of commercial PS beads with a blowing agent, such as isopentane. Steam heating of the impregnated beads leads to a cellular structure with a very low density. EPS is commercialized as beads or foams having widespread use in the packaging and building insulation sectors. High impact polystyrene (HIPS) is synthesized by emulsion polymerization of styrene in styrene-butadiene latex. The higher impact characteristics of HIPS make it suitable for use in the manufacture of sheets, food containers, window frames, household goods, etc.

Polyvinyl Chloride ( P V C )
PVC is a plastic of low crystallinity, prepared by polymerization of vinyl chloride at temperatures of about 50C. There are two main grades of PVC, rigid and flexible. Rigid PVC is the product directly obtained from the polymerization and, as its name indicates, it is a stiff, hard and often brittle polymer. Flexible PVC is obtained by blending with a variety of plasticizers, which leads to a soft and pliable material. Rigid PVC is used in the manufacture of sheets, pipes, window profiles, etc., whereas the applications of flexible PVC include wire coating, toys, floor coverings, films and tubing. In addition to plasticizers, PVC usually incorporates other components such as impact modifiers, fillers and extenders.

Polyethylene Terephthalate (PET)


Several routes are available for synthesis of PET, starting from different monomers: terephthalic acid (TPA), dimethylterephthalate (DMT) and bishydroxyethylterephthalate (BHET). The most common method of PET synthesis is based on the copolymerization of TPA and ethylene glycol. PET is a thermoplastic which can exist in amorphous, partially crystalline and highly crystalline states. For most PET applications, crystallinity is desired because it leads to enhanced strength and increases the maximum working temperature. PET is widely used in the manufacture of fibres, bottles and films. In recent years, rapid growth in the use of the moulding grades of PET has occurred.

Introduction

11

Thermosets
Thermosets are used in a similar proportion for both plastic and non-plastic applications. Plastic uses of thermosets include vehicle seats, sports equipment, electrical and electronic components, etc., while typical non-plastic applications include coatings and adhesives. The main commercial thermosets are urea-formaldehyde resins (UF), melamine-formaldehyde resins (MF), phenol-formaldehyde resins (PF), epoxy resins, unsaturated polyesters, alkyd resins and polyurethanes. Changes in thermoset consumption in Western Europe during the period 1994-1996 are shown in Table 1.2. UF/MF resins and polyurethanes are produced in the greatest quantities, making up about 70% of the total thermosets market.

Plastic Additives
Commercial plastics are typically prepared by mixing one or more polymers with a variety of additives in order to adjust and improve the properties and performance of the polymer. The main types of plastic additives are classified as follows:6 Plasticizers. The major role of plasticizers is to reduce the polymer modulus by lowering its glass transition temperature (Tg).Plasticizers are usually low molecular weight organic compounds having a Tg of about - 50 "C. Common plasticizers used for PVC include dialkyl phthalate, aliphatic diesters and trialkyl phosphate. Fillers and reinforcements. Fillers are inert materials used primarily to reduce the resin cost but also to improve the polymer processability. Typical fillers include clay, talc, silica, fly ash, mica, sand, glass beads, graphite and carbon black. Thermal stabilizers and antioxidants. These additives are used to protect the polymer against the effects of temperature and oxygen during processing. Free-radical scavengers such as hindered phenols and aromatic amines are typically added for this purpose. Light stabilizers. These are usually compounds which absorb ultraviolet light, in order to avoid degradation of the polymer by radiation. Products derived from benzophenone are typically used as light stabilizers. Flame retardants. These components are incorporated to inhibit or modify the polymer combustion when heated in an oxidative atmosphere. Typical flame retardants are Cl-, Br- and P-containing organic compounds, as well as antimony oxide and hydrated alumina. Colorants. Addition of soluble dyes and the dispersion of pigments are the methods used to provide plastics with desired colours. Dyes include azo compounds, anthraquinones, xanthenes and azines whereas a variety of inorganic compounds are used as pigments such as iron oxides, cadmium and titanium dioxide.

12
0

Chapter I

Antistatic agents. Static electrical charges may build up on the surface of polymers due to their low electrical conductivity, which may cause dust accumulation and sparking problems. These charges can be dissipated through the addition of external or internal antistatic agents (phosphate and fatty acid esters, sulfated waxes, quaternary ammonium compounds, amines, etc.). Blowing agents. These are used in the preparation of foamed plastics, mainly polystyrene and polyurethanes. Both physical and chemical blowing agents are used, that volatilize or decompose into gases after being mixed with the polymer. Examples of blowing agents include shortchain hydrocarbons (pentanes and hexanes), fluorocarbons, gases (nitrogen, air, carbon dioxide), hydrazine derivatives, etc.

The above paragraphs show the great variety of both organic and inorganic compounds that are added to polymers when manufacturing plastic goods. Moreover, the proportion of additives can be varied over a wide range, depending on the plastic application. Methods for the recycling or transformation of plastic wastes must take into account the possible presence of these additives.

Both natural and synthetic rubber are commercially used in the manufacture of a variety of goods. As mentioned earlier, rubbers are elastomeric polymers, characterized by the presence of a network structure that may be temporarily deformed when subjected to external forces. Natural rubber accounts for about 25% of total rubber consumption. It is produced from the Hevea brasiliensis tree, being formed by isoprene units with cis-1,4 links. Natural rubber is used in tyres and for retreading, latex, mechanical goods, etc. A variety of synthetic rubbers are commercially used: styrene-butadiene rubber (SBR), polybutadiene, ethylene-propylene rubber, butyl and halobutyl rubber, etc. The most important is SBR, which is mainly used as a major component of all passenger tyres and in significant amounts in most tyre products. Rubbers are usually subjected to a vulcanization or curing process to improve their properties. Vulcanization is carried out commonly by reaction with sulfur, which leads to the formation of a three-dimensional structure through the formation of sulfur bridges between the polymer chains. Other vulcanizing agents include peroxides, metal oxides, amines, etc. As in the case of plastics, rubber goods also incorporate a number of additives7
0

Accelerators of the curing process that allow control of the time and rate of vulcanization, as well as the number and type of sulfur crosslinks which are formed. Typical accelerators include guanidines, mercaptobenzothiazoles and sulfenamides, etc.

Introduction

13

Activators which increase the vulcanization rate by reacting with the accelerators yielding rubber-soluble complexes. Zinc oxide and stearic acid are widely used as activators. Retarders to delay the initial onset of vulcanization, providing the time necessary for processing the uncured rubber. Antioxidants and antiozonants. Process oils and plasticizers. Fillers. Carbon black, clays, calcium carbonate, silica, etc. are typical rubber fillers. Carbon black is used in relatively high proportions to improve the strength of the rubber. The performance of carbon black incorporated into rubber largely depends on its particle size distribution. Tyres contain over 30 wt% of carbon black.
The disposal of scrap tyres is currently an important environmental problem, as most used tyres are dumped in landfills. In some cases, accumulations of used tyres have accidentally caught fire, causing the release of toxic substances into the atmosphere. The generation of scrap tyres has been estimated to be of the order of 1.5 million tonnes per year in the European Union, 2.5 million tonnes per year in North America and 0.5 million tonnes per year in Japan.8

3 Plastic Wastes
The increase in the use of plastic materials in all sectors of industry and in everyday life, as well as the reduction in the lifetime of most plastic products, have led to a continuous increase in the generation of plastic wastes. Figure 1.7 illustrates the distribution of solid wastes by sector in Western Europe for 1996. The total volume of waste was 2.6 x lo9 tonnes, although the total post-user plastics waste accounted for only 1.7 x lo7 tonnes, i.e., only 0.6 wt% of the solid waste was plastic materials. Due to the extensive use of plastics in packaging, most of the plastics waste is found in domestic refuse. Table 1.3 summarizes the plastics waste generated by

OTHER WAST
30 70%

TRlBUTlON AND RGE INDUSTRY 12 50%

MUNICIPAL SOLID WASTE 5 00% AUTOMOTIVE 0 50% AGRICULTURE 3a 50%

Figure 1.7 Solid wastes by sector generated in Western Europe ( 1996) o 2

14

Chapter I

Table 1.3 Recovery o f plastics waste by sector in Western Europe for 1996 (based on reference 2 )
Sector Post -user plastics waste (ktonneslyear)
10 673 329 1

MSW Distribution and large industry Building/construction Electrical/electronics Automotive Agriculture Total

832 828 920


327

16 871

sectors in Western Europe in 1996. Approximately 63% of the total plastics waste is disposed of in municipal solid waste (MSW) mixed with a wide variety of other solid wastes. There is a relationship between the sector that originates the plastic waste and the life of the plastic products. Plastics used in packaging and agriculture generally have a lifetime of less than 1 year, hence in this case the composition of the waste directly reflects the resin consumed. Plastics used in the manufacture of household items and electrical devices usually have a lifetime of between 1 and 10 years, whereas the duration of plastics used in construction, automotive and furniture is typically over 10 years. Therefore, in the last two cases the wastes currently being generated account for products that were produced and consumed some time ago. The average composition of household waste in Western Europe is shown in Figure 1.8, the plastic content in MSW being just 8 wt%. A similar proportion is found in the USA, where in 1993 plastics made up 9.3 wt% of the total household waste. The percentages by weight of the different resins found in plastics household waste are shown in Figure 1.9. Polyolefins (L/LDPE, HDPE and PP) account for approximately 60% of the total plastics waste in MSW, which is in close correspondence with the production of these polymers. Likewise, over 90% of the total plastics waste is made up of just six resins (LDPE, HDPE, PP, PS, PVC and PET).
PLASTICS GLASS

RGANIC PRODUCTS 35% PAPEWCARDBOA 25%

Figure 1.8 Average composition of the M S W in Western Europe (1996).2

Introduction Others 9.2%


LlLDPE

15

PET 8.1Yo

PS I EPS 11.9%

19.6%

Figure 1.9 Plastic distribution by resin corresponding to M S W in Western Europe ( I 996).2

The Economic and Environmental Impact of Plastic Wastes


From an economic point of view, used plastic can be considered as both an important source of valuable chemicals, mainly hydrocarbons, and an energy source. The calorific value of most plastics is similar to that of fuel oils and higher than that of coals. Plastic wastes can therefore be viewed as potential fuels, when other alternatives of valorization are not possible.' Plastic wastes represent a significant environmental impact due to the following facts: Because of their resistance to degradation, plastic materials exist for a long time when disposed of in landfills. Although some biodegradable plastics have been successfully synthesized, most of the commercial resins are not found within this category. Moreover, the slow degradation of plastics is responsible for the progressive reduction of landfill capacity. The risk of accidental fires with highly polluting emissions is increased in landfills containing large amounts of plastics. Plastic wastes account for about 25% of all solid wastes accumulated in landfills. Plastics usually contain a variety of additives such as fillers, stabilizers, plasticizers, reinforcing agents, colorants, etc. Both organic and inorganic compounds are added to plastics to improve and modify their properties, containing in many cases heavy metals. Thus, according to the United States Environmental Protection Agency, plastics contribute 28 Yo of all cadmium present in MSW and about 2% of all lead." As a consequence of their low density, plastics cause a greater visual impact on disposal than many other materials. Similarly, the light weight of plastics is the origin of important limitations on the recycling of plastic wastes, due to increased collection and transportation costs. Thus, to

16

Chapter 1

recover 1 tonne of used plastic it is necessary to collect about 20000 plastic bottles. Because plastic wastes are found in MSW mixed with other solid wastes, complex and costly separation steps are required to produce used plastic streams of relatively high purity.

Despite all these problems, substitution of plastics by other materials is not environmentally sound. According to Gebauer and Hofmann, l 1 replacement of plastics in packaging by glass, paper, cardboard or metals would lead to drastic increases in the weight ( > 400%), cost ( >200%) and volume (>200%) of the packaging products, as well as in the energy consumed (>200%) in their manufacture. The advantages of using plastic materials are demonstrated by the following examples:
0

Energy requirements for polyethylene grocery sacks are 2 0 4 0 % lower than paper, while generating about 75% less solid waste, 65% less atmospheric emissions and 90% less waterborne waste. Replacement of about 200-300 kg of conventional materials in a modern car by plastics leads to a reduction in the fuel consumption of 750 litres over a lifespan of 150 000 km. If the whole automotive sector of Western Europe is considered, this reduction would cause a decrease in the oil consumption of 12 million tonnes and in CO2 emissions of 30 million tonnes every year.2 In the case of glass containers, 43% by volume of a lorry load would be the packaging, whereas if using plastic containers this value is reduced to 7Y0.l~

'*

Management of Plastic Wastes


The general concerns about environmental protection and resource conservation have led to the development of a variety of solid waste management techniques to reduce both the environmental impact of the different types of waste and the depletion of natural resources. Management of plastic wastes cannot be treated as an individual problem; it must be considered as an integral part of the global waste management system. Current waste management is based on a four-level hierarchical approach:
0

Reduction. Minimizing the consumption of raw materials through improvements in the design of products may allow a significant reduction in the amount of wastes generated when they reach the end of their life cycle. As examples of the progress in this area, Table 1.4 shows the decrease in the weight of different food containers over the period 19701990. However, it is clear that there is a limit to the advances which can be made by weight reduction, since the mechanical properties and performance of the products are also affected by this decrease. Reuse. This is mainly applied to packaging goods, being defined as any

Introduction

17

Table 1.4 Reduction in weight of food containers (1970-1990)


Container Wine bottle (glass) Beer bottle (glass) Supermarket bag (PE) Yoghurt container (PS) Coke bottle (PET) Weight in 1970 (8) Weight in 1990 (g) Reduction (%) 22 38 70 45 35

450 210 23 6.5 66

3 50 130 6.5 3.5 42

operation by which packaging items are refilled or used for the same purpose for which they were conceived, with or without the support of auxiliary products. l4 Consumers and industries are encouraged to promote the reuse of goods and packaging instead of disposal. This option can be applied especially for containers such as bottles, bags, etc. Recycling. This allows the wastes to be reintroduced into the consumption cycle, generally in secondary applications because in many cases the recycled products are of lower quality than the virgin ones. Recycling must be applied only when the amount of energy consumed in the recycling process is lower than the energy required for the production of new materials. Plastics can be recycled using two different approaches: mechanical and feedstock recycling. In the first case, the plastics are recycled as polymers, whereas in the second, plastic wastes are transformed into chemicals or fuels. Energy recovery. When the recycling of wastes is not feasible or there is no market for the recycled product, incineration can be used to generate energy from the waste combustion heat. Plastics are materials of high calorific value, hence plastic wastes greatly contribute to the energy produced in incineration plants. Alternatively, they can be used as fuels in a number of applications: power plants, industrial furnaces, cement kilns, etc. Incineration of Cl-containing plastics has been the subject of great controversy due to the possible formation and release into the atmosphere of dioxins. However, the relationship between PVC content in the waste stream and dioxin concentration has not been clearly demonstrated. In fact, it seems that the formation of dioxins depends mainly on the incineration conditions rather than on the waste composition.

On the basis of this hierarchical approach, only non-recyclable and nonenergy valuable wastes should be disposed of in landfills. However, the current picture of plastic wastes management is far from this situation. Figure 1.10 shows the destiny of plastic wastes in Western Europe during 1996. Most of the plastics are still disposed of by landfilling, followed by energy recovery and with small proportions of mechanical and chemical recycling. This distribution has been forecast to change in the next few years, with more recycling and energy recovery and a large decrease in the amount

18

20000

16000
n

8 t

5
v

12000

2
8000

I -

n
I -

0
4300

Figure 1.10 Plastic waste management in Western Europe (1996) .2

of plastic wastes sent to landfills. In fact a rapid growth in feedstock recycling has been observed in recent years in Europe, although this is concentrated mainly in one country, Germany. Targets for the minimization of wastes have been established by the EU Directive on Packaging and Packaging Waste, which came into force in December 1994, fixing minimum and maximum recovery rates of 50 and 60%, as well as minimum and maximum recycling targets of 25 and 45%. Moreover, a minimum recycling rate of 15% was established for each individual material. There are several reasons for the lack of recycling of plastic wastes, compared to other solid materials such as paper/cardboard and glass. A number of problems arise from the large variety of chemical compositions and properties of the different types of plastics, which makes it difficult to establish a general recycling procedure. In addition, plastic wastes are mainly contained in MSW, mixed with other solids. Therefore, costly and complex separation treatments must be applied in many cases to obtain a plastic waste stream having a more or less homogeneous composition. Likewise, the low density of most plastics makes it necessary to deal with large volumes of wastes in order to produce a given mass of recycled material.

Introduction

19

Mechanical Recycling
As shown in Figure 1.1 1, mechanical or material recycling of plastics involves a number of treatments and operations: separation of plastics by resin, washing to remove dirt and contaminants, grinding and crushing to reduce the plastic particle size, extrusion by heat and reprocessing into new plastic Because thermosets cannot be remoulded by the effect of heat, this type of recycling is mainly restricted to thermoplastics. Mechanical recycling is limited by the compatibility between the different types of polymers when mixed, as well as by the fact that the presence of small amounts of a given polymer dispersed in a matrix of a second polymer may dramatically change the properties of the latter, hindering its possible use in conventional applications. Thus, the presence of low amounts of PVC in recycled PET strongly reduces the commercial value of the latter, due to the possible release of HC1 during the PET reprocessing. This problem is enhanced by the fact that PVC and PET are difficult to separate from other plastic wastes. Another difficulty with mechanical recycling is the presence in plastic wastes of products made of the same resin but with different colours, which usually impart an undesirable grey colour to the recycled plastic.

Plastic Waste

Shredder

Vibrating ring section Cone press

Stationary cone

Pin mill Screen changer Pelletizer

Figure 1.11 Process for the mechanical recycling o f plastics.I5

20

Chapter I

In addition, most polymers suffer a certain degradation during their use due to the effect of a number of factors such as temperature, ultraviolet radiation, oxygen and ozone. This degradation leads to a progressive reduction in length and to a partial oxidation of the polymer chains. Therefore, recycled polymers usually exhibit lower properties and performance than the virgin material, and are useful only for undemanding applications. Recycling plastics without prior separation by resin produces a material with mechanical properties similar to timber, hence it is often used for the replacement of timber in certain applications. A higher quality of recycled plastics is achieved when separation by resin is carried out prior to the remoulding step. However, even in this case, recycled plastics cannot be used in food containers, unless direct content with the food can be avoided. An alternative developed in recent years for promoting the use of recycled plastics has been the preparation of containers with a three-layer wall. The middle layer is the thickest and is made of recycled polymer, whereas the thinner external and internal layers are made of virgin material. With this approach direct contact between the recycled polymer and both the consumer and the product in the container is avoided.

Feedstock Recycling
The severe limitations on the mechanical recycling of plastic wastes highlight the interest and potential of feedstock recycling, also called chemical or tertiary It is based on the decomposition of polymers by means of heat, chemical agents and catalysts to yield a variety of products ranging from the starting monomers to mixtures of compounds, mainly hydrocarbons, with possible applications as a source of chemicals or fuels. The products derived from the plastic decomposition exhibit properties and quality similar to those of their counterparts prepared by conventional methods. A wide variety of procedures and treatments have been investigated for the feedstock recycling of plastic and rubber wastes. For the purposes of this book, these methods have been classified into the following categories (Figure 1.12):
0

0 0
0

Chemical depolymerization by reaction with certain agents to yield the starting monomers. Gasification with oxygen and/or steam to produce synthesis gas. Thermal decomposition of the polymers by heating in an inert atmosphere. Catalytic cracking and reforming. The polymer chains are broken down by the effect of a catalyst, which promotes cleavage reactions. Hydrogenation. The polymer is degraded by the combined actions of heat, hydrogen and in many cases catalysts.

The progress and the current status of these alternative methods of feedstock recycling are described in the following chapters. At present, feedstock recycling is limited by the process economy rather than by technical reasons. Three main factors determine the profitability of these alternatives: the degree of separation

Introduction

:
21
PLASTICS AND RUBBER WASTES

Figure 1.12 Alternativesfor the feedstock recycling of plastic and rubber wastes.

required in the raw wastes, the value of the products obtained and the capital investment in the processing facilities. In most of the above methods, some pretreatments and separation operations must be carried out on the plastic wastes prior to feedstock recycling, which results in an increase in recycling costs. According to the separation steps required, the different methods of feedstock recycling can be ordered as follows: gasification < thermal treatments % hydrogenation < catalytic cracking < chemical depolymerization. Many of the projects on chemical recycling of waste plastics have failed in the past due to the relatively low price of the derived products. In recent years, there has been a trend towards the production of added value compounds such as olefinic gases, paraffins, activated carbons, etc. In general terms, the commercial value of the products obtained in the different treatments can be ordered as: thermal oils z synthesis gas < hydrogenation oils z catalytic oils < monomers. It is interesting to note that the required pretreatments and product value follow almost reverse orders. However, many other factors should be included for an adequate comparison of these alternatives: the possibility of carrying out the treatment in existing or new facilities, minimum size of the industrial plants needed to be profitable, required investment, plant location, etc. Hofmann and Gebauerlg have identified the main problems involved in several of the feedstock recycling methods: High capital expenditure, especially in hydrogenation plants. Lack of a regional plastic waste volume to support the continuous operation of large-scale plants. For gasification, a minimum capacity about 400 000-500 000 tonnes per year is necessary. Pyrolysis and hydrogenation lead to a wide range of end-products, which then have to be further upgraded and processed, mainly in refinery units. Several petrochemical companies have considered the possible feedstock recycling of plastic wastes in existing refinery facilities, which would avoid the need to invest and build new processing This alternative is based on the similarity of elementary composition between plastics and petroleum fractions. Moreover, taking into account the differences in production between plastics and the total of petroleum-derived products, plastic wastes could be incorpo-

22

Chapter I

rated into refinery streams in relatively low amounts. The main problem associated with this approach is the possible presence of undesired elements and compounds (Cl, N, metals, etc.) in the plastic wastes that would be introduced into the refinery. In such cases, even a small drop in yield or efficiency, when multiplied by hundreds of tonnes, would have dramatic effects on the refinery economy. Accordingly, plastic wastes should be intensively pretreated and conditioned prior to being added to petroleum fractions. Figure 1.13 shows the main refinery units that may be used in the processing of plastic wastes according to Hofmann and Gebauer (visbreaker, hydrogenation and gasification). Meszaros2 has also proposed the processing of plastic wastes in the cokers and catalytic cracking units of refineries. The economic viability of this alternative greatly depends on the proximity of the plastic waste generation points and the refineries, due to the high costs of the plastic transportation.

Sorting and Separation of Mixed Plastics


The previous sections point out the relationship between the purity of the raw plastic wastes and the quality of the products obtained by both mechanical and feedstock recycling. Therefore, removal of contaminants and separation of plastics by resin can be considered as treatments required before the plastic recycling is carried out. A variety of methods have been developed for the separation of plastics: manual, flotation, dissolution, spectroscopic identification, etc. Manual separation of plastics is carried out by personnel placed on either side of a conveyor transporting the plastic goods. Figure 1.14 illustrates a manual sorting facility with a number of operators specialized in the separation of different plastic containers.22 Three sorters initially remove HDPE goods, and two sorters are responsible for separating green and clear PET products. Another operator separates mixed color HDPE and the final three sorters are in charge of the recovery of PP and PVC bottles and any remaining PET and HDPE containers. As the incoming material stream progresses towards the end of the conveyor, plastic goods are progressively removed until nothing but trash and undesirable materials remain. Plastic separation by flotation is based on density differences between the polymers (see Table 1.5). Mixed plastics are dispersed in a liquid having a density between those of the different plastics. Polymers with densities lower than that of the medium float, while those of higher density sink, and so can be easily separated. In successive steps the density of the liquid medium can be changed in order to achieve separation of the resulting fractions. The density of aqueous solutions can be adjusted by adding salts or alcohols. In some cases the separation of a given polymer is promoted by adding a wetting agent and bubbling a gas through the solution, which selectively adheres to the polymer surface, causing it to float.23Likewise, addition of solvents that are absorbed by PVC, so reducing its density, has been proposed to favour the separation of PVC and PET by flotation.24 In a recent patent, the use of a solvent in supercritical conditions has been reported as an effective method for the

Introduction
Petroleum

23

VVA

Atmospheric distillation

I
Vacuum distillation

VVA

Mixing

Vi sR

E
Hydrogenation

f
Visbreaker
VVA
I

VVR

VisR
VvR/
\

Vacuum distillation
J

b
L

Mixing
1

Gasification

v
Synthesis gas ( CO, H ,)

Metha no1 synthesis

v
Methanol

Figure 1.13 Processing o f plastic wastes in rejnery units. VVA: pre-treated mixed plastic scrap; AR: residue from the atmospheric distillation process; VR: vacuum distillation residue; VisR: visbreaker residue; VVR: vacuum visbreaker residue.

separation of plastics.25 The extremely high compressibility of a fluid in the vicinity of its critical point allows its density to be varied by minimal changes in the temperature or pressure. Plastics with similar densities can thus be separated by adjusting the fluid density, usually using C02, to a value between those of the two polymers. Plastic separation by density differences can be accelerated by centrifugation instead of flotation techniques. Differences in solubility have also been exploited for the separation of

24

Chapter I

63

63

63

T I

Introduction

25

Table 1.5 Densities of various polymers (based on reference 3 )


Polymer Density (g/cm3)

HDPE
LDPE

PP PS PVC PET PMMA Nylon-6,6

0.94-0.97 0.91-0.93 0.93-0.94 1.04-1.10 1.39-1.40 1.33-1.39 1.19-1.20 1.20-1.30

plastics. A suitable combination of solvents and dissolution/precipitation steps allows the main plastics to be isolated. These methods usually involve temperature variations because the solubility of a polymer in a given solvent changes drastically with this variable. According to Nauman and Lynch,26a mixture of PS, LDPE, HDPE, PP, PVC and PET can be separated by fractional dissolution in xylene. As the solvent temperature is increased, progressive dissolution of these resins occurs as follows: PS (room temperature), LDPE (75"C), HDPE (105 "C), PVC (138 "C), whereas PET remains in the residue. Several spectroscopic techniques have been found to be useful in the identification and separation of plastic goods such as X-ray fluorescence, IR and ultraviolet spectroscopies, etc. IR-based methods are highly efficient, and have recently been used on a commercial scale. Figure 1.15 shows a facility for the separation of plastics via irradiation of the plastic objects with near-infrared radiation.27 The reflected radiation is collected and analysed in an IR detector, the resulting spectrum being assigned to a given polymer by comparison with a spectra library. All these steps take place within a few seconds as the object passes along a conveyor. Once the major resin has been identified, a mechanical or pneumatic system directs the plastic object to the conveyors of individual resins. Methods based on X-ray radiation operate in a similar way, the transmitted light spectrum from the object being compared with typical spectra of various resins using a computer. The performances of manual and automated sorting by X-ray radiation facilities for the separation of plastics by resin and colour on a commercial scale has recently been compared.22 The post-consumer plastic materials produced after sorting were: clear PET, green PET, two natural HDPE grades of different quality, mixed color HDPE, PP and PVC. The automated system was found to be more cost-effective, because additional sorting labour is necessary at the manual facility. The automated plants were also more capital intensive, although the difference in capital costs regarding the manual sorting was not enough to compensate for the labour savings in the automated facilities. With respect to separation efficiency, both types of facility produced high quality recycled resins. However, the manual system partially failed in the separation of PET and PVC objects, which led to a low value PET product. Even after years

26

Chapter 1

MIXED PLASTIC CONTAINERS

-.

UNIDENTIFIED OBJECTS

PET

PVC

PE

PP

PS

Figure 1.15 Automated sorting facility of plastic wastes based on IR radiation27: IR source (I), IR detector (2), computer with spectrum library ( 3 ) , main conveyor ( 4 ) , compressed air ( 5 ) , solenoid valves ( 6 ) ,pusher ( 7 ) .

of experience, operators in manual systems have problems in separating PVC and PET bottles. Taking into account the cost and difficulty of plastic separation, several initiatives have recently been adopted to make these operations easier. A new concept of recycling is gradually being introduced, based on rethinking the entire life cycle of a product, in order to develop products that have a lesser effect on the environment. If recycling is considered from the beginning of the design process, a number of actions could be taken to favour product recycling:28 Design for easy product disassembly, especially when using multiple materials in the same object. Likewise, the attachment of plastic and nonplastic parts should be avoided. Minimization of the number of resins commercially used. Instead of tailored polymers for very specific applications, the use of general and versatile resins should be emphasized. Try to avoid the use of a variety of polymers in the same object. If more than one type of plastic is used, it is important to ensure they are compatible, so they can be recycled together. Avoid the use of polybrominated biphenyls and diphenyls or other noxious compounds as plastic additives. Compliance with eco-label requirements. The introduction of codes for the identification of the major plastics can be a very efficient means of

Introduction

27

PET

H DPE

PVC

LDPE

PP

PS

Others

Figure 1.16

Codes for the identijcation of plastic objects by resin.

facilitating polymer separation. Figure 1.16 shows the code widely adopted for different resins, which should be marked on plastic goods.

Future Trends in Plastic Waste Management


While currently most plastic and rubber wastes are still disposed of in landfills, it is forecast that in the next few years the significance of other alternatives will have to be enhanced, as the number of landfill sites progressively decreases in many countries. There will have to be an increase in both mechanical and feedstock recycling and, when they are not feasible, energy recovery may be the most suitable option. At present, the different methods of plastic recycling are limited on a commercial scale by the process economy. The decrease in the price of virgin polymers which has occurred in recent years has negatively affected many recycling programmes, because the value of the recycled resins is fixed as a percentage of the virgin polymer price. Fluctuations in the price of virgin resins of 50% or even more within a year are quite usual. The development of recycling processes leading to higher quality products would help the recycling economy. In addition, standardization of the recycled resins may be an important factor for promoting their commercial application. The recycling policy in many countries has been focused in recent years on increasing the recycling rates with little emphasis on the search for applications for the recycled products. A number of measures have been suggested to promote plastic recycling? Standards requiring that certain products contain a minimum amount of recycled resin. Minimum recycling rates per sector and/or material. Procurement programmes that establish the obligation of purchasing items containing recycled materials. Both positive and negative incentives, such as fees imposed on resins that do not meet the minimum standards for recycling or tax credits to businesses using recycled resins in their products. Likewise, in the case of feedstock recycling, many processes could be economically viable if a

28

Chapter I

credit of a fixed amount was paid for every tonne of plastic waste that is diverted from landfill to recycling. Requirements for manufacturers to maintain responsibility for the final disposal of their products.

Recently, a survey was carried out in Western Europe of the possible evolution of plastic waste management during the period 1995 to 2006. The main conclusions of this report are as follows: Plastic wastes are predicted to increase by 6.6% per year. PET and PP wastes will be generated even faster, as a consequence of their higher consumption. Mechanical recycling has the potential to double, from 1.2 million tonnes in 1995 to 2.7 million tonnes by 2006, which corresponds to an average growth of 8.4% per year. The main constraints for improving mechanical recycling rates are: the imbalance between the collectable waste and the potential end-markets for recycled resins, and the presence of large quantities of mixed plastic wastes which are difficult to clean and sort. Other technologies such as feedstock recycling must be promoted to meet the recycling targets. This survey confirms that the solution to the problem of plastic wastes should be based on an integrated approach through the application of mechanical recycling, feedstock recycling and energy recovery processes. Of these, feedstock recycling presents a high potential for growth in the coming years provided that economically feasible processes can be developed.

References

1 APME, Assessing the Potential for Post-use Plastics Waste Recycling - Predicting Recovery in 2001 and 2006, Summary Report, APME, Brussels, 1998. 2 SOFRES Conseil for APME, Plastics. A Material of Choice for the 21st Century, APME, Brussels, 1998. 3 R. Gomez and J.R. Gil, Los Plasticos y el Tratamiento de sus Residuos, UNED, Madrid, 1997. 4 SOFRES Conseil for APME, Information System on Plastic Waste Management in Western Europe, European Overview, 1995 Data, APME, Brussels, 1997. 5 C. Hadjilambrinos, Environ. Conserv., 1996,23(4), 298. 6 J.R. Fried, Polymer Science and Technology, Prentice Hall PTR, Englewood Cliffs, New Jersey, 1995. 7 C.S.L. Baker and W.S. Fulton, in Ullmans Encyclopedia of Industrial Chemistry, Vol. 21, ed. W. Gerhartz, B. Elvers, M. Ravenscroft, J.F. Rounsaville, and G. Schulz, VCH Verlagsgesellschaft, Weinheim, 1993, p. 48 1. 8 C. Roy and J. Unsworth, in Pyrolysis and Gasification, ed. G.L. Ferrero, K. Maniatis, A. Buekens, and A.V. Bridgwater, Elsevier Applied Science, London, 1989. 9 F. Dawans, Rev. I. Fr. Petrol., 1992,47(6), 837. 10 EPA, Plastics: The Facts about Production, Use, and Disposal. From EPAs Report

Introduction

29

to Congress on Methods to Manage and Control Plastic Waste', United States Environmental Protection Agency, Cincinnati, 1990. 11 M. Gebauer and U. Hofmann, Proceedings of Recycle '93, Davos, Switzerland, 1993. 12 R.J. Rowatt, Chemtech, 1993,23(1), 56. 13 J. Singer, Resources Conserv. & Recycling, 1995, 14, 133. 14 S. Cervera-March, Proceedings of the 7th Mediterranean Congress on Chemical Engineering, Barcelona, 1996. 15 S. Moore, Mod. Plast. Int., 1993,23(2), 12. 16 S. Babinchak, Chemtech, 1991,21(12), 728. 17 M. Gebauer, KunststoJffe, 1995,85, 2. 18 R.D. Leaversuch, Mod. Plast. Int., 1991,21(7), 26. 19 U. Hofmann and M. Gebauer, Kunststofe German Plast., 1993,83,4. 20 M. Gebauer, Proceedings of Recycle '94, Davos, Switzerland, 1994. 21 M.W. Meszaros, Proceedings of Recyclingplas VI '91, Washington, 1991, p. 335. 22 J. Burgiel, W. Butcher, R. Halpern, D. Oliver, P. Tangora, R.W. Beck, and D.R. Kirk, 'Cost Evaluation of Automated and Manual Post-consumer Bottle Sorting Systems', United States Environmental Protection Agency, Cincinnati, 1994. 23 K. Saitoh and S. Izumi, US Patent 4 132 633, 1979. 24 M.J. Grimm and T.R. Sehlmeyer, US Patent 4 167 111, 1986. 25 E.J. Beckman, US Patent 5 126 058, 1992. 26 E.B. Nauman and J.C. Lynch, US Patent 5 198 471, 1993. 27 H.D. Ruhl, Jr. and K.R. Beebe, US Patent 5 134 291, 1992. 28 G.L. Nelson, Chemtech, 1995,25(12), 50.

CHAPTER 2

Chemical Depolymevization
1 Introduction
The alternative methods of plastic recycling described in this chapter consist of the breakdown of the polymer by reaction with certain chemical agents, leading back to the starting monomers. These monomers are identical to those used in the preparation of virgin polymers, hence the plastics prepared from both depolymerization and fresh monomers are expected to have similar properties and quality. According to this approach, plastic wastes are reintroduced to the market as polymers, as happens in the case of material recycling, but without the loss of resin properties typically associated with the latter process. Plastic recycling by chemical depolymerization is the most established method of plastic chemical recycling. Different chemolysis processes have been applied on an industrial scale for several years.'92 The major disadvantage of chemical depolymerization is that it is almost completely restricted to the recycling of condensation polymers, and is of no use for the decomposition of most addition polymers, which are the main components of the plastic waste stream. Condensation polymers are obtained by the random reaction of two molecules, which may be monomers, oligomers or higher molecular weight intermediates, which proceeds with the liberation of a small molecule as the chain bonds are formed. Chemical depolymerization takes place by promoting the reverse reaction of the polymer formation, usually through the reaction of those small molecules with the polymeric chains. Several resins widely used on a commercial scale are based on condensation polymers, such as polyesters, polyamides, polyacetals, polycarbonates, etc. However, these polymers account for less than 15% of the total plastic wastes (see Chapter 1). Depending on the chemical agent used to break down the polymer, different depolymerization routes can be envisaged: glycolysis, methanolysis, hydrolysis, ammonolysis, etc. In the following sections of this chapter, these alternatives are reviewed for those condensation polymers having the most significant commercial applications. It must be pointed out that a majority of the studies on chemical depolymerization of plastic wastes is reported in patents; works published in the scientific literature are relatively scarce.
31

32

Chapter 2

2 Polyesters
Chemical depolymerization of polyesters has been mainly applied to polyethylene terephthalate (PET), the most common polyester on the market. Chemolysis of PET by a variety of methods has been known for many years. In fact, the chemical depolymerization of PET can be considered the starting point of plastic chemical recycling. PET is a semi-crystalline thermoplastic polymer used in the manufacture of fibres, packaging films, bottles, electrical insulators, etc. As shown in Scheme 2.1, PET can be produced by two different routes: by condensation between terephthalic acid (TPA) and ethylene glycol (EG) or through the reaction of dimethyl terephthalate (DMT) with ethylene glycol. Both alternatives lead to the monomer bis(hydroxyethy1) terephthalate (BHET), which is further polymerized into PET.

0
HO -CH2 -CH2TPA

0-E O ! @ BH ET

-CHZ-CH2

-OH

CHJ-O-!~!-O -C H ~

/
Scheme 2.1
Routes of P E T preparation.

- n H,O

DMT

PET

Different PET chemolysis methods have been developed aimed at the production of TPA, DMT or BHET, all of them being possible monomers for the reconstruction of fresh polyesters. The exact monomer formed by PET depolymerization depends on the type of chemical agent used to break down the polymeric chains. In certain processes, the final product of PET chemolysis is a mixture of polyols, useful in the formulation of other polymers such as unsaturated polyesters, polyurethanes and polyisocyanurates. This is an interesting case of chemical recycling because the breakdown of one polymer leads to the raw materials for the preparation of a quite different class of plastics. Depending on the depolymerization agent, polyester chemolysis methods have been classified as follows: glycolysis, methanolysis, hydrolysis, ammonolysis, aminolysis and combined processes. Figure 2.1 summarizes the different alternatives for PET chemolysis, as well as the type of products derived from each one. All these PET degradation alternatives are reviewed in the following sections.

Chemical Depolymerization

33

TPA

Polyols

Polycondensation

Diisocyanate
Hydrolysis

Polyurethane

Repolymerization 4

Methanolysis

Ammonolysis

TPA Amide

Figure 2.1 Main PET chemolysis alternatives.

Glycoly sis
Glycolysis is the simplest and oldest method of PET depolymerization. The first The method patents on PET glycolysis were filed more than 30 years involves the reaction of PET, under pressure and at temperatures in the range 180-240 "C, with an excess of glycol, usually ethylene glycol, which promotes the formation of BHET.1c13 This monomer has to be purified, normally by melt filtration under pressure, prior to its use in the production of new PET polymer. Colours present in the starting PET wastes are not usually removed by glycolysis. The depolymerization is carried out in the presence of a transesterification catalyst, usually zinc or lithium acetate. The following phenomena have been proposed to describe the mechanism of PET glycolysis:1 4 * 1 5 glycol diffusion into the polymer, polymer swelling which increases the diffusion rate, and reaction of the glycol with an ester bond in the chain. Because the reaction rate is proportional to the polymer surface area, it is advisable to first reduce the size of the raw PET waste to small particles by grinding, cutting, etc. Malik and Mosti6described a process for the depolymerization of PET with ethylene glycol catalysed by sodium acetate, with several improvements over previous methods. PET is first mixed with a certain amount of BHET to promote its dissolution; and the formation of diethylene glycol during the reaction is inhibited by continuously introducing water into the reaction mixture. The production of glycol ethers, mainly diethylene glycol, is undesir-

34

Chapter 2

able because they tend to copolymerize with BHET, forming polyesters of lower quality. Almost no excess of ethylene glycol is used because, although the depolymerization reaction would proceed more rapidly, it may promote the formation of diethylene glycol. The formation of diethylene glycol can be limited by using a combination of lithium and zinc acetates and antimony trioxide as ~ata1yst.l~ In addition to diethylene glycol, other secondary products have been identified during PET glycolysis, such as ethylene terephthalate, dioxane, aldehydes and cyclic trimers. l 8 The effect of the catalysts used in the PET chemolysis by reaction with ethylene glycol has been studied by Baliga and Wong.'' In this work the PET glycolysis was carried out at 190C with an excess of ethylene glycol (EG/PET = 1.3 weight ratio) in the presence of different metal acetates as catalysts: zinc, lead, cobalt and manganese acetates, which are typically used in transesterification reactions. In a blank experiment without catalyst, a considerable amount of PET was still detected after 8 h of reaction, showing that the presence of a suitable catalyst is a very important factor in promoting the depolymerization with reasonable times. In the catalytic experiments, the glycolysed products were formed by a mixture of BHET and some oligomers that did not change significantly with the type of catalyst. However, the type of metal acetate influenced the initial rate of depolymerization, the following order being observed: Zn2+ > Pb2+ > Mn2+ > Co2+. The glycolysis of both clear and green PET, recycled from post-consumer soft drink bottles, was also investigated in this work. Figure 2.2 shows the change in the hydroxyl number of the products with reaction time, which is a measurement of the extent of depolymerization. The hydroxyl number corresponding to BHET can be calculated as 441.5 mg KOH/g. From Figure 2.2, only slight variations between the glycolysis of both types of PET are observed, showing that the pigment present in the green recycled PET does not significantly affect the depolymerization rate. However, it tends to discolour the glycolysed products, so that further purification steps are necessary in order to obtain colourless BHET. The kinetics of PET degradation with ethylene glycol were investigated by Campanelli et al.13 A good fit of the results was obtained with a first-order kinetic model in both ethylene glycol and ethylene diester concentrations. Although zinc salts appeared to have a catalytic effect on glycolysis below 245"C, they do not influence the reaction rate above this temperature. Because PET melts at about 245"C, the authors proposed that the effectiveness of zinc compounds in PET chemolysis is related to interfacial phenomena, which are not present when the glycolysis reaction takes place in a single liquid phase. In addition to the possible reuse of BHET in the preparation of fresh PET polymer, two other processes have been proposed starting from the PET glycolysis products: preparation of unsaturated polyester^^&^^ and synthesis of polyester p o l y o l ~ . ~ The ~-~ latter ~ can be used in the formulation of polyurethanes and polyisocyanurate foams2c30 (see Figure 2.1).

Chemical Depolymerization
490.
390
h

35

, "
I C

380

370
ac E

360
L .

350
E Y

3h0

330
L . W

190 C .
0

clear recycled P E T green recycled P E T

.&

320

310

300

10

Glycolysis time (hour)

Figure 2.2 Hydroxyl number evolution during the glycolysis o f clear and green recycled PET. l 9 (From S . Baliga and W.T. Wong, J . Polym. Sci. Polym. Chem., 1989, 27, 2071. Reprinted with permission from John Wiley & Sons Ltd.)

A variety of glycols other than ethylene glycol have also been used to promote PET degradation. Thus, Vaidya and Nadkarni26327have investigated PET depolymerization with propylene glycol at 200C in the presence of zinc acetate as catalyst. The reaction product was a mixture of several monomers, dimers and trimers. The major fraction consisted of the following monomers: bis(hydroxypropy1) terephthalate, bis(hydroxyethy1) terephthalate and h ydr oxyprop y 1-h ydrox ye t h yl terep h thalate. Thereafter, these products were used in the synthesis of unsaturated polyesters by a polyesterification reaction with maleic anhydride at 180-2OO0C, followed by mixing with styrene monomer. Polyesterification is a reversible process involving the reaction between a diacid and a diol to yield a polyester and water. Using this method, PET waste can be converted into unsaturated polyesters, which are commercial products with an important added value due to their use in a number of applications, such as a matrix for fibre-reinforced composites. The main properties of the unsaturated resins synthesized from the products of PET glycolysis were determined and compared to those of conventional unsaturated polyesters. Both types of resin were found to have similar processability, although the PET-based unsaturated resins were more suitable for applications such as hot moulding, where a higher viscosity is desirable. The effect of the type of glycol used in the chemical depolymerization of PET has been studied at 200 "C, comparing the results obtained with ethylene glycol,

36

Chapter 2

propylene glycol and diethylene glycol.31Likewise, the effect of the glycol/PET ratio was investigated using ethylene glycol.3oThe highest extent of depolymerization was observed with ethylene glycol, which was related to its higher molar concentration, because the experiments were conducted at the same glycol percentage by weight. In these studies, the glycolysis products were polyesterified by reaction with adipic acid at 170-200 "C to yield polyester polyols, which were then reacted with 4,4'-diphenylmethane diisocyanate to yield both polyurethane elastomers and rigid polyurethane foams, demonstrating that PET wastes can also be recycled into this class of polymers with important industrial applications. This process has been further commercialized to produce polyols that can meet different application requirements by changing the type and proportion of the glycol in the PET transesterification step.33The contamination level of the initial PET wastes is not a limiting factor, although if they are dark in colour, the obtained polyurethane and polyisocyanurate foams are also dark, which decreases their commercial value. A wide range of chemical agents, catalysts and conditions for the glycolysis of unsaturated polyester resins, used in the manufacture of buttons, have been described in a recent patent.34 In addition to different metal acetates, the following compounds have been proposed to be catalytically active in PET glycolysis: sodium methylate, sodium ethylate, sodium hydroxide, methanesulfonic acid, magnesium oxide, barium oxide and calcium oxide. Different applications of the depolymerization products were described, e.g., preparation of fresh unsaturated polyesters by reaction with maleic acid, maleic acid/ phthalic anhydride or maleic anhydride/terephthalic acid or the synthesis of polyurethane resins by reaction with a diisocyanate. Lusinchi et al.35have explored the recycling of PET + PVC mixtures through glycolysis of the polyester component. Treatment of PET at 190C with ethylene glycol in the presence of zinc acetate led to 80% BHET and 20% dimer. After removal of the dimer, BHET was polycondensed with c-caprolactone and the resulting oligomers reacted with hexamethylene diisocyanate to yield a variety of polyurethane polymers. Increasing the E-caprolactone content in the co-oligomer allowed polyurethanes which were miscible with PVC to be obtained. These results show the feasibility of PET + PVC recycling through its transformation into PU + PVC blends. Moreover, the mechanical properties of the blend could be adjusted by varying the chemical structure of the co-oligomer and the amount of PU. Other glycolytic methods for PET depolymerization described in the patent literature involve reaction with an alkene oxide,36 mainly ethylene and propylene oxides, at temperatures between 120 and 160 "C.The reaction was catalysed by basic compounds: sodium hydroxide, potassium hydroxide and tertiary amino alkyl phenols. The polyol mixture obtained had a hydroxyl number in the range 140-240. These polyols were blended with conventional polyols and the mixture used in the preparation of polyurethane and polyisocyanurate foams, which had better fire resistance than foams made only with conventional polyols. An alternative glycolytic method of PET degradation is based on treatment at

Chemical Depolymerization

37

220 "C with a digesting medium comprising a mixture of phthalic anhydride and diethylene The resulting polyol-rich products are said to be useful in the preparation of polyurethane and polyisocyanurate foams. Finally, an interesting process for PET recycling by glycolysis has recently been patented.38 It involves the introduction of polyester waste into a mixture containing dimethyl terephthalate and ethylene glycol, in which the polyester is first subjected to depolymerization followed by a polycondensation step using the same reaction mixture. The two steps take place at different temperatures and with different catalysts: 200-210C in the presence of metal acetates for depolymerization and 280C in the presence of antimony trioxide for polycondensation. The significance of the process lies in the solubility of PET waste in DMT/EG mixtures, which allows PET to be depolymerized and repolymerized, without any separation steps, to obtain new polyester resins containing up to 75% of the used constituents. Interestingly, the colour of the polyester product is not affected by the addition of this high proportion of PET wastes.

Methanolysis
PET methanolysis is based on the treatment of PET with methanol at relatively high temperatures (180-280 "C) and pressures (20-40 atm), which leads to the formation of dimethyl terephthalate (DMT) and ethylene glycol as the main The reaction proceeds usually in the presence of typical transesterification catalysts, the most widely used being zinc acetate. Other catalysts employed in PET methanolysis are magnesium acetate, cobalt acetate and lead dioxide. Originally, PET methanolysis was developed by PET manufacturers as a process aimed at the recovery and treatment of polyester wastes generated during the production cycle in order to increase the polyester yield. However, with the increase in environmental public concern, methanolysis began to be considered as a feasible alternative for the recycling of PET residues present in the solid waste stream. The products of the methanolysis are usually separated and purified by distillation or crystallization. Purified DMT can be reintroduced into the PET polymerization process with properties similar to those of virgin DMT. Compared to BHET, the monomer obtained by PET glycolysis, DMT is produced by methanolysis with a higher purity in regard to physical contaminants. Nevertheless, some organic impurities cannot be completely removed, which may cause some poor colour. A number of methods for PET methanolysis have been described in the patent literature, operating under both batch and continuous conditions. One of the major problems with the continuous process is the difficulty of introducing the solid polyester wastes into the methanolysis reactor working under high pressure. For this reason, methanolysis often takes place in batch systems, with all the problems and limitations associated with batch operation. The reaction temperature can be reached by heating and melting the PET wastes in a first step, which are subsequently contacted with methanol. In other

38

Chapter 2

methods the heat necessary to melt the PET polymer is supplied by contact with superheated methanol v a p ~ u r s,45947 . ~ ~In this case, methanol acts as both a heat transfer agent and a chemolysis agent. In some cases, solvents are used to facilitate contact between the two reagents participating in the methanolysis reaction. This method was used by Naujokas and Ryan,45 who proposed a methanolysis process in which scrap PET waste is dissolved in oligomers of DMT and ethylene glycol, and superheated methanol is passed through the resulting solution. Recently, PET methanolysis has been carried out with supercritical methanol at a temperature of 300 "C and pressures above 80 atm.48 Under these conditions, PET decomposition was much faster than when using liquid methanol, leading to the production of DMT and some oligomers.

Hydrolysis
Reaction of PET with water allows the polyester chains to be broken down into terephthalic acid (TPA) and ethylene glycol. The process can be carried out under neutral, acidic or basic conditions. A crucial aspect of this chemical recycling method is the purity and properties of the obtained TPA in order to achieve the specifications normally required for direct esterification to produce fresh PET polymer. TPA is usually purified by crystallization from solvents such as acetic acid, whereas a variety of procedures have been reported to remove the different impurities present in the hydrolysis product. Several patents have been filed dealing with the acid hydrolysis of PET by reaction with concentrated sulfuric acid ( > 14.5 M).49-51The process takes place at temperatures between 25 and 100C with a duration of just a few minutes at atmospheric pressure. The hydrolysis product is treated with sodium hydroxide to neutralize the TPA produced, which causes the formation of the corresponding TPA sodium salt, which is soluble in water. The solution obtained is usually of a dark colour, but this can be removed by ion-exchange columns. In the final stage of the process, the resulting solution is again acidified to reprecipitate TPA, which is obtained with a purity > 99%. One of the major drawbacks of this process is the corrosion induced by the reaction mixture and the formation of large amounts of liquid wastes, containing inorganic salts and sulfuric acid, which must be disposed of. The effect of sulfuric acid concentration on the acid hydrolysis of PET has been investigated by Yoshioka et al.52who tried to develop a process using less concentrated sulfuric acid. As shown in Figure 2.3, the degradation of PET increases slightly with acid concentrations up to 5 M. Beyond this point, a sharp increase in the PET hydrolysis is observed. At a sulfuric acid concentration of around 7 M, PET degradation is almost total. These results indicate that it is possible to achieve PET hydrolysis with dilute sulfuric acid. However, it is necessary to work at higher temperatures (1 50 "C)and for extended periods of time (up to 5 h) compared to the conventional PET acid hydrolysis processes. When studying the effect of time, these authors also observed a significant increase in PET conversion after 2 h of reaction. From SEM micrographs of the

Chemical Depolymerization
100

39

80

60

40

20

H,SO, concentration/ M

Figure 2 . 3 Eflect of the sulfuric acid concentration in PET hydrolysis (150 "C, 5 h): 0 PET conversion, A TPA yield, 0ethylene glycol yield.52

degraded PET powder, this abrupt change was related to an increase in the PET specific surface area through the formation and growth of cracks on the polymer surface. The alkaline hydrolysis of PET involves treating the polyester with an aqueous solution of sodium hydroxide (4-20 wt%) under pressure at temperatures between 200 and 250C for periods of several Under these conditions the sodium salt of TPA is formed and by acidification TPA is recovered from the solution as a precipitate. It has been observed that the rate of the PET alkaline hydrolysis increases in the presence of quaternary ammonium compounds. Thus, Niu et al? have reported on the alkaline degradation of PET fibres with addition of dodecylbenzyldimethylammonium chloride (DBDMAC) into the reaction mixture. A sharp increase in the PET hydrolytic degradation at 80 "C was observed with DBDMAC concentrations in the range 0-1.0 g/l, especially for the least crystalline fibres. The authors concluded that the rate enhancement by quaternary ammonium compounds occurs preferentially on the amorphous regions of the PET fibres. In recent years, several works have appeared on the alkali degradation of PET in solvents other than water. Thus, Collins and Z e r ~ n i a n have ~ ~ reported that treatment of PET with methanolic sodium hydroxide causes a faster degradation than when using aqueous sodium hydroxide. In the same way, Oku et al?' converted PET quantitatively by reaction at 150 "C with sodium hydroxide dissolved in anhydrous ethylene glycol. Disodium terephthalate was obtained as the major product, being precipitated from the ethylene glycol solution. Terephthalic acid was easily recovered by dissolution in water of the sodium salt followed by acidification with HC1. 'H-NMR measurements of the

40

Chapter 2

obtained TPA indicated its high purity, with no ethylene glycol being detected. In a further the authors showed that non-aqueous alkaline PET degradation can also be carried out efficiently by reaction with potassium hydroxide dissolved in ethanol or methanol. Moreover, the incorporation of different ethers as co-solvents resulted in enhanced PET conversions. Figure 2.4 shows the variation in yield of the dipotassium TPA salt (TPA-K2) with reaction time when PET is degraded at 50 "C with KOH dissolved in mixtures of ethanol with dioxane, tetrahydrofuran and 1,2-dimethoxyethane. The greatest increase in the PET degradation rate is achieved with the addition of 20% dioxane as cosolvent, which allows yields of over 90% to be achieved in just 30 min, in spite of the low reaction temperature. A promising PET chemolysis method is neutral hydrolysis based on treatment with liquid water or steam. In contrast with the acid and alkaline hydrolysis, PET is degraded without the formation of unwanted liquid effluents containing inorganic salts. Moreover, although some decrease in the pH occurs due to TPA formation, the reaction system is not corrosive, and so standard materials can be used in the apparatus. Neutral hydrolysis is usually performed under pressure (10-40 atm) at temperatures in the range 200-280 0C.59-62 Alkali metal acetates are typically used as transesterification catalysts to promote PET hydrolysis. The reaction proceeds more slowly than the acid hydrolysis, several hours being required to achieve high PET conversions. When the hydrolysis is

I .o

08

06

04

02

00

20

40

60

80

Reaction time (min)

Figure 2.4

TPA-K2 yield versus time during PET decomposition by treatment at 50 "C with KOH in a rnixed solvent of ethanol and ethers (80/20 ~01%): 0no ether, 0 dioxane, A tetrahydrofuran, I ,2-dirneth0xyethane.'~

Chemical Depolymerization

41

carried out with steam, the steam acts as a major source of heat for the hydrolysis zone, it stirs up the waste material so accelerating the hydrolysis reaction, and its partial condensation provides the liquid water necessary for the reaction. The effect of zinc catalysts on PET hydrolysis has been investigated by Campanelli et a1.62in the temperature range 250-280 "C. The catalytic effect of zinc salts was attributed to the electrolytic changes induced in the polymerwater interface during hydrolysis. Likewise, Kao et al.63concluded the existence of an autocatalytic mechanism in PET hydrolysis because the depolymerization reaction is catalysed by the carboxyl groups produced during the reaction. One of the major problems associated with the neutral hydrolysis method is that most of the impurities initially present in the PET waste remain in the TPA produced by the reaction. Therefore, complex and intensive purification operations are needed to obtain TPA with properties similar to the commercial grades. A hydrogenation step has been proposed as a method for the removal of impurities and colour found in the TPA produced by PET neutral hydrolysis.60 TPA precipitated from the hydrolysis medium is slurried in water and catalytically hydrogenated at 260-290 "C and a pressure of 65-82 atm for around 1 h. Palladium supported on carbon is one of the preferred catalysts for this hydrogenation step. This treatment leads to a colour level and fluorescence properties in the produced TPA similar to those of the commercially available virgin PET. Hydrolytic treatments can serve not only as a PET degradation method, but may simultaneously enable the separation of hydrolysable and non-hydrolysable polymers present in the plastic waste stream. Thus, Saleh and well ma^^^^ have proposed the separation of PET and polyolefin mixtures by treatment with water from about 200C up to the critical temperature of water under autogenous pressure. The resulting liquid phase contains the hydrolysis products, TPA and ethylene glycol, whereas the solid phase is formed by the nonreacted polyolefins.

Ammonolysis and Aminolysis


These treatments have been less widely investigated for PET chemical recycling. Ammonolysis consists of the reaction of PET with ammonia at temperatures between 70 and 180 "C, usually under pressure and in the presence of ethylene The main product of the degradation is the TPA amide, which is obtained with a purity above 99% and in yields of around 90%. This reaction is also catalysed by zinc acetate, as in other PET chemolysis processes mentioned above. PET depolymerization by aminolysis is based on the reaction with primary amines such as methylamine, ethylamine and e t h a n ~ l a m i n eThe . ~ ~degrada~~ tion takes place at temperatures in the range 20-100 "C and the studies carried out show that the attack of the amine is located preferentially on the amorphous regions of the polymer, because it is difficult to achieve a total PET conversion.

42

Chapter 2

The final products of PET aminolysis are the amides of TPA and ethylene glycol.

Combined Chemolysis Methods


In recent years new processes for PET chemical recycling have been patented that cannot be classified in any of the previous sections, because they use more than a chemical agent to promote the polyester cleavage. Usually, these methods consist of two or more steps which combine different types of treatments: glycolysis-hydrolysis, methanolysis-hydrolysis, glycolysismethanolysis, etc. The major goal of these combined treatments is to benefit from the advantages of each individual process. Doerr7 has reported the development of a two-step process (glycolysishydrolysis) for the chemical recycling of PET and other condensation polymers. In the first step, PET is reacted with ethylene glycol at 280-290C in a highpressure screw extruder with a residence time of 2 min in order to reduce its molecular weight by at least 50%. The resultant products are rich in bis(2hydroxyethyl) terephthalate (BHET) and a variety of other hydroxyl-terminated oligomers. The second step of the process involves the neutral hydrolysis of these products by reaction with water, which leads to the formation of TPA and ethylene glycol as final products. The major advantage of this combined process is the decrease in the reaction time of the hydrolysis step compared to the hydrolytic processing of PET that has not been pre-glycolysed. Figure 2.5 shows the changes in PET molecular weight with reaction time for different treatments. Whereas the conventional neutral hydrolysis requires 45 min to yield the monomers, this is reduced to just 15 min in the two-step process. Another advantage of this method is the smaller hydrolysis reactor needed, and therefore the lower investment necessary for this unit. Moreover, this process has been applied to other polyesters, such as polybutylene terephthalate. In this case, the depolymerizing agent used in the preglycolysis step is 1,4-butanediol. Interest in the sequential combination of glycolytic and hydrolytic treatments for the chemical recycling of PET is demonstrated by the fact that some semiindustrial processes have recently been developed based on this approach, an example being the process developed by Smorgon Consolidated, an Australian producer of PET bottles. The first stage involves washing the waste with hot water in order to separate paper labels and polyolefins using a float/sink hydrocyclone. After drying, the washed flake is passed through a reactor where it is treated with ethylene glycol at 197C to partially depolymerize the PET component into a brittle state. After glycolysis, the product is passed between two steel rollers to break the embrittled PET into small pieces, which makes possible its separation from the other non-embrittled melted plastics, such as polyamides, polyolefins and PVC. In a subsequent step, PET depolymerization is completed by hydrolysis at 200C to yield TPA, which is insoluble under these conditions, and so easily separated by decanting. Similarly, the combination of methanolysis and hydrolysis has also been proposed as an interesting alternative to obtain fibre grade TPA from PET

Chemical Depolymerization
20,000 30,oOc

43

PET
MW

REACTION TIME

TO COMPLETION

(MINI

Figure 2.5

treatment^.^' (0 M.L. Doerr, U S Patent, 4 620 032, 1986)

Change o f the PET molecular weight along the time f o r diflerent degradation

wastes.72The initial treatment consists of the reaction of PET with superheated methanol vapours at 240-260 "C under atmospheric pressure, which causes the PET to depolymerize into its constituent monomers, such as dimethyl terephthalate, monomethyl terephthalate and ethylene glycol, as well as oligomeric products having a degree of polymerization in the range 5-20. The monomers are continuously removed from the reactor together with the methanol vapours, whereas the oligomers remain in the reactor to complete their degradation. The methanolysis products are fractionated by distillation, the column bottom being fed into the hydrolysis reactor working at a temperature around 270 "C with an excess of liquid water. The TPA acid produced is separated by precipitation, whereas the contaminants and insoluble materials present in the starting waste remain in the mother liquors. According to the authors, the process is effective even when the starting PET is very contaminated by materials such as metals, dyes, paper, films and other resins. Finally, a third possible combination of treatments (glycolysis-methanolysis) for PET recycling has been reported by Gamble et al.73In this case, the first step

44

Chapter 2

is the treatment and dissolution of PET at 230-290 "C with a mixture formed by ethylene glycol, terephthalic acid and dimethyl terephthalate oligomers. This stream is recycled from the final melt product of the process, which eliminates the need for an external source of reactants, resulting in a cost-effective and continuous recovery method. The product of the glycolytic treatment is further subjected to methanolysis by contact with superheated methanol at temperatures between 250 and 290C, which leads to a mixture rich in DMT. Most contaminants, as well as polyolefinic plastics that may be present in the PET waste, are accumulated in the first reactor, which enables the production of high purity DMT.

Comparison of the Various PET Chemolysis Methods


In this section the alternative methods of PET chemical recycling described earlier are critically compared and examined with a view to their possible largescale application. A number of industrial plants carrying out PET degradation are currently in operation, based mainly on methanolysis and glycolysis treatments. Hydrolytic processes are less advanced, most of them being used at laboratory and pilot-plant scales, although several projects are being developed to be applied commercially in the next few years. Ammonolysis and aminolysis based processes are rather less well established and developed treatments, hence their industrial application is not anticipated in the near future. Gebauer2 has compiled valuable information on the different industrial and semi-industrial plants for PET chemical recycling currently in existence. The largest capacities for the treatment of PET wastes are in methanolysis plants. Paszun and S p y ~ h ahave j ~ ~ compared the different PET chemolysis methods in regard to factors such as feed flexibility, reaction conditions, safety, economics, etc. Likewise, matt hew^'^ has evaluated the energy consumption of the main PET chemical recycling methods. These studies show that hydrolytic processes are advantageous, with an energy consumption between 20 and 30 MJ kg-' of PET, compared with the energy requirements in methanolysis and glycolysis of 40-60 MJ kgWith respect to feed specifications, methanolysis is the least flexible process, requiring well-defined industrial wastes. Glycolysis and hydrolysis may be able to manage more contaminated polyester wastes. Strong depolymerization conditions are used in methanolysis and hydrolysis, especially in acid and alkaline hydrolysis. In methanolysis, the presence of superheated methanol necessitates stringent safety conditions, whereas in acid and alkaline hydrolysis the reaction medium is highly corrosive. Moreover, the latter processes generate liquid effluents with soluble inorganic salts, which must be disposed of, causing additional environmental problems. On the contrary, neutral hydrolysis proceeds without these drawbacks, although the PET degradation rate is significantly lower. From the point of view of product versatility, glycolysis is particularly interesting because, depending on the reaction conditions, different oligomer mixtures can be obtained with potential applications not only in the

'.

Chemica1 D epolymer izat ion

45

preparation of fresh PET, but also in the synthesis of unsaturated polyester resins and in the formulation of polyurethane and polyisocyanurate foams. Finally, from an economic point of view, methanolysis is a very capitalintensive process, and hence it is economically interesting only for large productions. This is not the case for glycolysis, which can also be profitable for small and medium plants. This is the reason for the large size of most methanolysis plants, with capacities up to 100 000 tonnes/year. No reliable data are yet available on the economics of the hydrolytic process. It must be pointed out that two-step combined processes have great potential for future industrial application. Glycolysis-hydrolysis treatments are especially interesting because they combine most of the individual advantages of the two alternatives: weaker degradation conditions in the glycolysis step which decreases the reaction effort necessary in the hydrolysis unit. Moreover, a large quantity of the contaminants present in the starting PET waste are separated after the first treatment, which favours product purification. The final products are TPA and ethylene glycol, which can be easily repolymerized into PET.

3 Polyurethanes
Polyurethanes are usually classified as engineering polymers characterized by the presence of the carbamate group (-0-CO-NH-). The major route for the preparation of polyurethanes is the reaction between a diisocyanate and a hydroxyl-rich compound with at least two hydroxyl groups, according to Scheme 2.2,

O=C=N-R-N=C=O
Diisocyanate

HO-R"OH
Diol

___)

$8C-NH-R-NH-C-0-R-0 8
Polyurethane

Scheme 2.2 Polyurethane synthesis by polymerization of a diisocyanate and a diol.

where R can be an aliphatic, cycloaliphatic or polycyclic group. Of the different diisocyanates used, the most common are methylene-4,4'-diphenyldiisocyanate, toluene-2,4-diisocyanate and hexamethylene diisocyanate. Likewise, low molecular weight hydroxyl-terminated polyesters or polyethers are typically employed as dihydroxyl compounds. The main properties of polyurethanes are their high strength and abrasion resistance, and they are inert against oxygen and ozone. Most of the polyurethanes produced are used in the formulation of rigid and flexible foams and elastomers, with a wide variety of applications as adhesives, coatings, insulating agents, etc. Foams are obtained by blowing air through the polymer or by adding water which reacts with residual isocyanate groups. The degree of rigidity of the resulting polymer can be controlled by changing the R and R' groups, which may contain additional hydroxyl groups leading to a crosslinked polymeric matrix. Polyurethane foams are extensively used in the automotive

46

Chapter 2

industry due to their contribution to improvements in safety, comfort, part integration and weight reduction of vehicles. With regard to the chemical recycling of polyurethanes, two aspects must be highlighted. Polyurethanes are used to manufacture durable goods, which means that it takes several years for these items to be disposed of into the solid waste stream. Moreover, polyurethanes contain around 4 wt% of N, which may hinder their recycling by oxidative treatments such as incineration or gasification due to the potential release of significant amounts of NO, in the gaseous effluents. The most important chemolysis methods so far developed to reverse the polyurethane polymerization reaction shown in Scheme 2.2 are glycolysis and hydrolysis. These processes are reviewed next, together with other less widely investigated treatments.

G1ycolysis
A variety of processes for polyurethane degradation by reaction with different glycols has been described in the literature during the last 30 Polyurethane glycolysis is usually carried out with an excess of glycols at temperatures around 200C and in many cases working at atmospheric pressure. After several hours of reaction, the polyurethane is completely liquefied and depolymerized, and catalysts are not necessary. The chemistry of the glycolytic reaction has been described by Ulrich et a1.81782 It involves the transesterification of the carbamate group by addition and reaction with the glycol. These authors verified this scheme by reacting benzyl phenylcarbamate and ethylene glycol at 195 "C, a total conversion of the former being observed after 3 h according to the reaction shown in Scheme 2.3. Because water-blown polyurethane foams contain diarylurea linkages, due to the reaction of diisocyanates with water, they also investigated the glycolysis of N,N-diphenylurea as a model compound. The results obtained clearly showed that the urea linkages are also glycolysed in these reaction conditions, indicating

Benzyl-phenyl-carbamate

2-Hydroxyethylphenylcarbamate
V
h

NH-C-0-NH
Diphenylurea

II

Scheme 2 . 3 Glycolysis of benzyl-phenyl-carbamate and diphenylur.ea.82

Chemical Depolymerization

47

that the polyols derived from water-blown polyurethane foams will present amino end groups. In the initial works on polyurethane glycolysis the reaction product was separated into two liquid phases,77978 which is not desirable because it hinders the reuse of the polyols. This problem can be avoided by using a mixture of glycols.83 Thus, treatment of scrap polyurethane foam with a mixture comprising 90-95% of a glycol (ethylene glycol, diethylene glycol, propylene glycol or dipropylene glycol) and 5-10% of ethanolamine has been reported to lead to a homogeneous single polyol-containing phase. Another important factor related to polyol reuse is the control of the final product viscosity, which can be achieved by varying the scrap foam/glycol ratio or by changing the type of glycol used. Glycolysis has been successfully applied to rigid and flexible polyurethane foams, as well as to polyisocyanurate foam. The polyols obtained by degradation are blended with virgin polyols in a proportion of around 50%.8' These mixtures can be used in the formulation of new rigid and semi-rigid foams, which exhibit properties similar to those of completely virgin foams. However, the use of 100% polyurethane degradation polyols is limited by their lower reactivity and higher viscosity compared to virgin polyols. In a recent patent,84 glycolysis has been reported to be a feasible chemical recycling method for the degradation of polyurethane/polyurea or polyurea wastes. Conventional alcoholysis of these polymeric wastes leads to products with a high content of urea groups and low molecular weight primary aromatic amines, which limits their application in the isocyanate polymerization reaction because the amine groups are much more reactive towards isocyanates than the alcohol groups of the polyols. These problems have been solved by a two-stage process: reaction of the polymeric wastes at 200C with a diol or polyol (ethylene glycol, diethylene glycol, hexanediol, glycerol, etc.) followed by treatment of the alcoholysis products with urea or a carbamic acid ester. It has been found that this second reaction greatly reduces the amine content, so that the final product is suitable for reuse in the isocyanate addition polymerization. Today, several industrial plants based on glycolytic treatment are in operation for the chemical recycling of polyurethane wastes, mainly those generated from the insulation and automotive sectors2

Hydrolysis
Hydrolysis is the second most important method of chemical recycling of polyurethanes. Various studies have been published dealing with polyurethane degradation by reaction with liquid water (150-200C) or steam (200320 0C).85-90 The hydrolytic reaction proceeds as shown in Scheme 2.4.

Polyurethane

Diamine

Scheme 2.4

Polyurethane hydrolysis.

48

Chapter 2

Polyols, diamines and carbon dioxide are the final products formed by polyurethane hydrolysis. The reaction between the diamine and phosgene allows the corresponding isocyanate to be formed, whereas the subsequent polymerization of this isocyanate and the polyols yields the starting polyurethane again. Mahoney et al.87 have described the reaction of polyurethane foam and superheated water at 200 "C for 15 min, which leads to toluene diamines and polypropylene oxide. Hydrolysis of polyurethane and rubber mixtures has been used as a method not only of recovering valuable chemicals from the polyurethane fraction, but also to separate the polymers because rubber is inert to hydr~lysis.'~ The degradation takes place by contact with saturated steam at 200 "C for 12 h. This process may find particular applications in the treatment of rubber/polyurethane laminations. The mechanism and kinetics of the reaction of polyurethane foam with dry steam have been investigated by Gerlock et al.91using a polyether-based toluene diisocyanate as starting material. Reaction with steam at temperatures between 190 and 250 "C caused the destruction of all urea and urethane linkages to yield a high quality polyol product, although a significant difference in reactivity between these two types of linkages was observed. The authors propose that the urethane bonds are rapidly broken by hydrolysis according to the reaction shown in Scheme 2.5, whereas the urea linkages undergo a slow thermal dissociation to form the corresponding isocyanate and amine. Finally, the formed isocyanate is also hydrolysed, increasing the yield of toluene diamine (TDI).

Polyurethane ( polyether-toluenediisocyanate)

Diisocyanate

TDA

*
TDA

NH

~A'';~~~-NH--'&H~

+ co2

Scheme 2.5

Mechanism for the hydrolysis of a polyether-based toluene diisocy~nate.~'

The effect of basic catalysts on these reactions was also investigated, a sharp acceleration being observed with the addition of sodium hydroxide at levels of

Chemical Depolymerization

49

about 2.9 mg per 100 mg of foam. The polyols obtained by this treatment were mixed with virgin polyol in a ratio of 20/80, yielding a high quality flexible foam.

Ammonolysis and Aminolysis


Polyurethane chemolysis by reaction with ammonia or various amines has been described in the literature. Sheratte92 has proposed a process based on polyurethane decomposition by various agents. Several examples were provided for polyurethane degradation with ethanolamine (120 "C), ammonia and ammonium hydroxide (180 "C), diethylene triamine (200 "C) and other basic reagents. In all cases the process involves, simultaneously or subsequently, reaction with propylene oxide, which allows the different amines obtained to be quantitatively converted into polyols according to the reaction shown in Scheme 2.6. The polyols derived from this process were used in the reformulation of new polyurethanes by polymerization with the corresponding isocyanate, and were suitable for application in rigid foams.

R-NH

2 CHJ-CH-CH
3
Propylene oxide

CH2-CH-CH
R-N:

PH
OH

3 3

CHz-FH-CH
Polyol

Amine

Scheme 2.6 Conversion of primary amines into polyols by reaction with propylene oxide.'*

An interesting ammonolysis process has recently been developed93 based on the reaction of polyurethane with ammonia under supercritical conditions, which favours both the degradation reactions and separation of the polyols produced. The flow diagram of this process is shown in Figure 2.6. Two different polyurethanes were used as starting materials: a solid elastomer based on a trifunctional polyethertriol, 1,4-butanediol and methylenebis(pheny1 isocyanate); and a flexible foam where the diol was replaced by water. The ammonolysis reactions were carried out at 139 "C and 140 atm for I20 min, and with a polyurethane/ammonia weight ratio of 1. Under these conditions the polyurethane conversion was practically total. The ammonolysis reaction transforms the CO group into urea and the ester groups and derivatives of carboxylic acids into amides, whereas ether and hydroxy groups are inert towards ammonia. Scheme 2.7 illustrates the stoichiometry proposed by the authors for the ammonolysis of the polyether urethane. After the reaction, urea is separated by extraction with water whereas the polyol remains as a residue in the reactor. Therefore, under supercritical conditions the polyether polyols are separated from the mixture at the same time that the ammonolysis reaction progresses. The diamines and the diol can be separated by distillation or precipitation. The phosgenation of the amine leads to the corresponding diisocyanate, which together with the polyol and the diol may be used in the recovery of the raw polyurethane.

50

Chapter 2

Aqu. solution
I

1. Ammono

Separation of Diol
4

1
Phosgena tion of Amine

Figure 2 . 6 Flow diagram of a process for polyurethane ammonolysis by treatment with supercritical ammonia.93

Polyethcrurethane

I
5 H 2 N e C H 2 e N H 2 +
Diamine

20 NH,,

4 OHfCH2fOH
Did

10 H2N-C-NH

HOW
Polyether polyol

Urea

Scheme 2.7 Stoichiometry of the ammonolysis of a polyether p re thane.^'

Combined Chemolysis Methods


Similarly to the case of PET chemical recycling, several promising alternatives have been described for polyurethane chemical recycling through a combination of treatments. The Ford hydroglycolyis process is a good example of these combined alternatives because it couples hydrolytic and glycolytic reactions to degrade the polyurethane chains.94 This process was developed whilst trying to solve some of the problems present in conventional PU glycolysis, in which a complex mixture of products is obtained, made up of diamines, aminocarbamates, urealinked aminocarbamate oligomers, glycols and polyols. The composition is

Chemical Depolymeriza tion

51

difficult to predict, hence the properties of the polyurethane produced using this recycled mixture directly are difficult to control. An alternative is the separation and purification of the produced polyol, but this is not an easy task. On the contrary, the product obtained in the hydroglycolysis process is made up of significantly fewer products, which makes purification of the polyols more feasible. Figure 2.7 is a flow diagram of the Ford hydroglycolysis process. In the reactor, polyurethane foam is degraded in the presence of water, diethylene glycol and alkali metal hydroxides at 200C. When NaOH is added to the reaction mixture as a catalyst, a cleaner polyol is obtained due to the absence of carbamates and ureas in the product - they are transformed into amines and alcohols by hydrolysis. After 4 h of reaction, all the polyurethane is decomposed, yielding polyols, amines, isocyanates and carbon dioxide. Separation of the products takes place by extraction with hydrocarbons such as hexadecanes (separation tank l), which dissolve most of the polyols. Moreover, phase separation occurs at about 160C and the polyols are expelled from the hydrocarbons (separation tank 2), being purified by vacuum distillation. Finally, the glycol-rich phase, coming out of the first settling tank, is filtered to remove impurities and also vacuum distilled to separate the amines and diethylene glycol, the latter being recycled back to the reactor. Variations of

WASTE FOAM

AMINES

VACUUM STILL

WASTE

Figure 2.7

Flow diagram o f the Ford hydroglycolysis process for the degradation o f

polyurethane^.^^
(Reprinted with permission from J. Braslaw and J.L. Gerlock, Znd. Eng. Chem. Process Des. Dev., 23,552. 0 1984 ACS)

52

Chapter 2

this process have been patented by isolation of the glycolytic and hydrolytic treatments in different steps and/or the use of steam as hydrolysing agent.95*96 The quality of the polyols produced in this hydroglycolysis process was evaluated by using them in the formulation of new polyurethanes by partial replacement of commercial polyols. The foam so produced did not have significantly different physical properties than the all-virgin foam, even with replacement levels of 50%. The combination of glycolysis and aminolysis has also been described in the ' literature for the chemical recycling of polyurethanes. Kondo et ~ 1 . ~have reported the degradation of both polyurethane and polyisocyanurate foams by reaction with a mixture of a glycol (diethylene glycol, dipropylene glycol, 1,4butanediol or 1,Spentanediol) and monoethanolamine at around 200 "C for 30 min. The liquid product obtained was used in the production of rigid polyurethane foam that, although slightly brown in colour, exhibited properties suitable for application as a heat insulating material. Similarly, a recent patent combines aminolysis and hydrolysis reactions for achieving polyurethane d e c o m p ~ s i t i o nThus, . ~ ~ scrap polyurethane is reacted with a mixture of diethanolamine and aqueous sodium hydroxide. The simultaneous attack of these agents on the polymeric chains allows the reaction time to be appreciably shortened. The reaction product, obtained as an emulsion, is subjected to a second treatment with propylene oxide in order to transform the amines and ureas present in the mixture into polyols, giving a final product which is substantially free of any hydrogen-containing nitrogen atoms. The polyols produced have been found to be particularly suitable for the preparation of fresh polyurethane polymer which can be used as an elastomer or flexible foam.

Polyamides, Polycarbonates and Polyacetals


This section describes the chemical recycling, via chemolysis, of certain condensation polymers which, although being produced in significantly lower amounts than polyesters and polyurethanes, are used in important applications, and so also contribute to the plastic waste stream. Polyamides are obtained either by the condensation of a dicarboxylic acid and an alkylene diamine or by the head-to-tail condensation between an amino carboxylic acid or the corresponding lactam. Polyamides may have aliphatic or aromatic chain backbones. Aliphatic polyamides (nylon) have the most important commercial applications, mainly in the manufacture of fibres. Nylon-6 and nylon-6,6 account for around 85% of all nylon currently used. Nylon-6 is derived from the polymerization of E-caprolactam, whereas nylon-6,6 is obtained by the condensation of hexamethylene diamine and adipic acid. At present, the principal method of polyamide chemical recycling is hydrolysis, which can be carried out under neutral, acidic or basic conditions. Several commercial processes are already operating with this technology.* Acid hydrolysis by contact with superheated steam has been mainly applied

Chemical Depolymerization

53

to nylon-6 As shown in Scheme 2.8, the reaction of this polymer with water causes the cleavage of the C-N bond, leading to the formation of the corresponding amino carboxylic acid, which under acid conditions is transformed by dehydration into ecaprolactam, the starting monomer of nylon-6. Of the possible acid catalysts, H3P04 and H3B03 are typically used at temperatures in the range 250-350 "C. Compared with the alkali-catalysed process, acid hydrolysis of nylon-6 leads to the production of purer Ecaprolactam, which makes its further recovery and purification easier. In the case of nylon-6,6, basic hydrolysis is the preferred treatment. Depolymerization occurs by reaction with sodium hydroxide, hexamethylene diamine and sodium adipate being the initially formed products. In a second step, the latter is converted into adipic acid via acidification with HC1.
NYLON-6 0 0 0 -NHfCH2fC-NH{CH2$C-NH-fCH2fCII II II
5 5

Aminocarboxylic acid

1
...........

Caprolactam ......................................................................... ..................................................................................... NY LON-6,6 0 0

1
Sodium Adipate

NaOH
+ NH2fCH21fNH2

NaOOCfCH2tCOONa

Hexamethylene diamine

HCI
+ 2 NaCl

HOOCfCH21f4COOH Adipic acid

Scheme 2.8 Hydrolysis o f nylon-6 and nyl0n-6,6.~

Ammonolysis has also been studied as a method of chemically recycling nylon. McKinney l o ' has disclosed the treatment of nylon-6,6 and nylon-6,6 plus nylon-6 mixtures with ammonia at temperatures between 300 and 350C and pressure of about 68 atm in the presence of ammonium phosphate catalyst to

54

Chapter 2

yield a mixture of monomeric products: hexamethylene diamine, adiponitrile and 5-cyanovaleramide from nylon-6,6, and E-caprolactam, 6-aminocapronitrile and 6-aminocaproamide from nylon-6 (Scheme 2.9). The ammonia proportion is at least one equivalent of ammonia per amide group. The reaction also produces water which, according to the reaction equilibrium, inhibits the conversion of the amides into nitriles. This equilibrium can be shifted by continuous removal of water from the reaction medium as it is formed. In a subsequent step most of the monomers can be transformed into hexamethylene diamine by hydrogenation. In a further patent, McKinney'02 has reported enhanced conversions by the use of Lewis acid catalysts such as Zn, Co, Pd, Mn, Ti and W chlorides.
NYLON-6

Aminocaproamide

Aminocapronitrile NY LON-6,6

Hexamethylene diamine

Adipamide

Adiponitrile

Scheme 2 . 9 Mechanism of the ammonolysis of nylon-6 and nylon-6,6.lo2

The mechanism and kinetics of polyamide ammonolysis have been studied by Kalfas.lo3 According to the reaction shown in Scheme 2.9, polyamide ammonolysis proceeds by amide link cleavage and amide end dehydration reactions, as well as ring addition and ring opening reactions for cyclic lactams in nylon-6. Cleavage of the amide links by ammonia causes the scission of the polymer into two shorter chains, resulting in the formation of an amine and an amide end group. Dehydration of the latter leads to nitrile end groups. Hydrolysis has been the main method used for the chemical recycling of other condensation polymers, such as polyacetals and polycarbonates. Hydrolysis of polyacetals leads back to the starting monomers, formaldehyde or trioxane. Polycarbonates are polymers synthesized by the reaction of phosgene and a dihydric phenol, commonly bisphenol A. Chemical recycling of polycarbonate

Chemical Depolymerization

55

wastes by treatment with aqueous sodium hydroxide has been proposed as a method for the recovery of Na2C03and bisphenol A.'04 Recently, treatment of polycarbonate with catalytic amounts of sodium hydroxide in a mixed solvent of methanol and toluene or dioxane at 60C has been reported to yield bisphenol A in solid form and dimethyl carbonate in solution. lo5

5 Summary
Chemical recycling of plastics by chemolysis is a well-established method of recovering the raw monomers and synthesizing fresh polymers with properties and applications similar to those of virgin resins. However, in most cases the products of the polymer degradation must be subjected to intensive purification operations to completely remove the various contaminants present in the plastic wastes. Several commercial processes are currently in operation for the chemical depolymerization of polyesters, polyurethanes and polyamides. These polymers are broken down by reaction with a number of chemical agents: water or steam (hydrolysis), glycols (glycolysis), methanol (methanolysis), ammonia (ammonolysis), amines (aminolysis), etc. In some cases, supercritical conditions have been used to promote the depolymerization reactions. PET glycolysis, mainly by reaction with ethylene glycol, leads to the formation of BHET, which can be used as a monomer for the preparation of fresh PET or, alternatively, it can be utilized in the formulation of unsaturated polyesters and polyurethane and polyisocyanurate foams through the corresponding polyesterification reactions. Treatment of PET with methanol yields DMT, which can also be repolymerized into PET. Compared to PET glycolysis into BHET, DMT is obtained with a higher purity by methanolysis, although it is a very capital-intensive process. PET hydrolysis takes place by reaction with liquid water or steam, leading to the formation of terephthalic acid. The process can be carried out under acid, basic or neutral conditions. Although stronger conditions are necessary to promote PET depolymerization by neutral hydrolysis, this a preferred alternative because large amounts of aqueous salt solutions are produced in the acid and alkali catalysed degradations. Other less well-established PET chemolysis methods include degradation by ammonolysis and aminolysis. Degradation of PET by combined treatments has evolved in recent years as an interesting alternative, benefiting from the advantages of each individual process. Thus, a number of combined processes have recently been developed: glycolysis-hydrolysis, methanolysis-hydrolysis and glycolysis-methanolysis. Polyurethane chemolysis can be performed by processes similar to those applied to PET. Thus, polyurethane glycolysis yields a mixture of polyols, which can be reused in the formulation of new polyurethanes. Likewise, polyurethane hydrolysis leads to the formation of polyols, diamines and carbon dioxide. The diamine can be subsequently transformed into the corresponding isocyanate by reaction with phosgene, whereas the polymer-

56

Chapter 2

ization of the polyols and the isocyanate yields the polyurethane. Ammonolysis processes have also been developed for the degradation of polyurethane wastes. In this way, treatment with supercritical ammonia allows more rapid depolymerization, while it also favours the separation of the formed urea, polyols, diol and diamine. Combined processes have also been developed to promote polyurethane chemolysis: hydrolysis-glycolysis, glycolysis-aminolysis and amino1ysis-hydrolysis. Chemical depolymerization of pol yamides is mainly carried out by hydrolysis. Acid hydrolysis of nylon-6 allows the starting monomer, &-caprolactam,to be recovered. Likewise, basic hydrolysis of nylon-6,6 leads to hexamethylene diamine and adipic acid. Degradation of polyamides by ammonolysis has also been reported as an interesting alternative for the chemical recycling of nylon-6 and nylon-6,6 mixtures.

6 References
1 2 3 4 5 6 7 8 9 10 11 12
13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 S. Shelley, K. Fouhy, and S. Moore, Chem. Eng., 1992, July, 30. M. Gebauer, Chem. Tech., 1996,48,2. P. Parrini and P. Parisini, Ital. Patent 702065, 1966. F. Siclari, D. Ruta, and F. Cauzzi, Ital. Patent 729 187, 1966. J. Heinze, H. Ramm, and H. Richardt, Ger. Patent 1 247 29 1, 1967. K. Miura, Y. Kagiya, and T. Ichikawa, Jpn. Patent 68 23449, 1968. H. Grushke and M. Mayer, Ger. Patent 1290929, 1969. G.M. Stevenson, Ger. Patent 1 803 929, 1969. Y. Etienne and R. Soulas, Fr. Patent 1 563 765, 1969. A. Fujita, M. Sato, and M. Murakami, US Patent 4609 680, 1986. J.Y. Chen, Y.C. Ou, and C.C. Lin, J . Appf. Polym. Sci., 1991,42, 1501. P.L. Johnson and D. Teeters, Pofym. Prepr. ( A m . Chem. SOC., Div. Polym. Chem.), 1991, 32, 144. J.R. Campanelli, M.R. Kamal, and D.G. Cooper, J . Appf. Pofym. Sci.,1994, 54, 1731. A.A. Petrov and E.M. Aizenshtein, Khimicheski, 1977,3, 30. A.A. Petrov, L.P. Repina, and E.M. Aizenshtein, Khimicheski, 1984,5, 8. A.I. Malik and E.E. Most, US Patent 4078 143, 1978. K.T. Barkey, US Patent 3 830 759, 1974. D. Gintis, Makromof. Chem., Macromol. Symp., 1992,57, 185. S. Baliga and W.T. Wong, J . Polym. Sci. A , 1989,27, 2071. H. Toshima, Jpn. Patent 75 64382, 1975. 0. Makimura and H. Miyake, Jpn. Patent 78 09275, 1978. S.N. Tong, D.S. Chen, C.C. Chen, and L.Z. Chung, Polymer, 1983,24,471. K.S. Rebeiz, D.W. Flower, and D.R. Paul, J . Appf. Polym. Sci., 1992,44, 1649. K.S. Rebeiz, Diss. Abstr. Int. B, 1993, 53,4269. T.P. Pepper, US Patent 5 380793, 1995. U.R. Vaidya and V.M. Nadkarni, Znd. Eng. Chem. Res., 1987,26, 194. U.R. Vaidya and V.M. Nadkarni, J . Appl. Pofym. Sci., 1987,34,235. L.L. Chandler, T.R. Friar, and B.W. Green, US Patent 4546 169, 1985. W.L. Carlston, R.T. Stoehr, and G.R. Svoboda, Mod. Pfast., 1985,62, 100. U.R. Vaidya and V.M. Nadkarni, J . Appl. Pofym. Sci., 1988,35,775. U.R. Vaidya and V.M. Nadkarni, J . Appf. Pofym. Sci.,1989,38, 1179. G. Tersac, G. Laurencon, and H. Roussel, Caoutch. Pfast., 1991,68, 81.

Chemical Depolymerization

57

33 A. De Leon, in Chemical Aspects of Plastics Recycling, ed. W. Hoyle and D.R. Karsa, The Royal Society of Chemistry, Cambridge, 1997, p. 86. 34 S. Kubota, 0. Ito, and H. Miyamoto, Eur. Patent 0 693 527 A l , 1996. 35 J.M. Lusinchi, Y. Pietrasanta, J.J. Robin, and B. Boutevin, J. Appl. Polym. Sci., 1998, 69, 657. 36 M.E. Brennan, US Patent 4439 549, 1984. 37 R.K. Hallmark, M.J. Skowronski, and W.D. Stephens, US Patent 4 873 268, 1989. 38 B.J. Subblett and G.W. Connell, US Patent 5 559 159, 1996. 39 S.A. Yeboah, Process. Contr. Qual., 1997, 9, 189. 40 R. Lotz, G. Wick, and C. Neuhaus, US Patent 3 321 510,1967. 41 H. Gruschke, W. Hammerschick, and W. Medem, US Patent 3 403 115, 1968. 42 K.T. Barkey, E.B. Lefferts, and D.C. May, US Patent 3 488 298,1970. 43 F. Ligorati, G . Aglietti, and V.E. Nova, US Patent 3 776 945, 1973. 44 M.N. Marathe, D.A. Dabholkar, and M.K. Jain, GB Patent 2041 916,1980. 45 A.A. Naujokas and K.M. Ryan, US Patent 5 051 528, 1991. 46 P. Muhs, A. Plage, and H.D. Schumann, Kunststofe, 1992,82,289. 47 B.R. Debruin, A.A. Naujokas, and W.J. Gamble, US Patent 5 432 203, 1995. 48 T. Sako, T. Sugeta, K. Otake, N. Nakazawa, M. Sato, K. Namiki, and M. Tsugumi, J. Chem. Eng. Jpn, 1997,30(2), 342. 49 G.E. Brown, Jr and R.C. OBrien, US Patent 3 952 053, 1976. 50 S.F. Pustaszeri, US Patent 4 355 175, 1982. 51 N.D. Sharma, A.A. Vaidya, and P. Sharma, Ind. Patent 163 385,1985. 52 T. Yoshioka, T. Sat0 and A. Okuwaki, J. Appl. Polym. Sci., 1994,52, 1353. 53 J. Pitat, V. Holcik, and M.A. Bacak, GB Patent 822 834, 1959. 54 J. Benzaria, B. Durif-Varambon, F. Dawans, and J.B. Gaillard, Eur. Patent 597 751, 1994. 55 S. Niu, T. Wakida, S. Ogasawara, H. Fujimatsu, and S. Takekoshi, Textile Res. J., 1995,65(12), 771. 56 M.J. Collins and S.H. Zeronian, J. Appl. Polym. Sci.,1992,45, 797. 1997,63,595. 57 A. Oku, L.-C. Hu, and E. Yamada, J. Appl. Polym. Sci., 58 L.-C. Hu, A. Oku, E. Yamada, and K. Tomari, Polym. J., 1997,29(9), 708. 59 J.W. Mandoki, US Patent 4605 762, 1986. 60 B.I. Rosen, US Patent 5095 145, 1992. 61 D.J. Royal1 and J.L. Harvie, Eur. Patent 550979, 1993. 62 J.R. Campanelli, D.G. Cooper, and M.R. Kamal, J. Appl. Polym. Sci., 1994, 53, 985. 63 C.-Y. Kao, B.-Z. Wan, and W.-H. Cheng, Ind. Eng. Chem. Res., 1998,37, 1228. 64 R.Y. Saleh and W.E. Wellman, US Patent 5 326 791, 1994. 65 K.P. Blackmon, D.W. Fox, and S.J. Shafer, Eur. Patent 365 842, 1988. 66 M.S. Ellison, L.D. Fisher, K.W. Alger, and S.H. Zeronian, J. Appl. Polym. Sci., 1982, 27, 247. 67 V.A. Popoola, J. Appl. Polym. Sci., 1988,36, 1677. 68 M.J. Collins, S.H. Zeronian, and M.L. Marshall, J. Makromol. Sci.,Chem., 1991, A28, 775. 69 Y.W. Awodi, A. Johnson, R.H. Peters, and A.V. Popoola, J. Appl. Polym. Sci., 1987,33,2503. 70 M.L. Doerr, US Patent 4620032, 1986. 71 Eur. Plast. News, 1994, April, 14. 72 S. Contessa and R. Vosa, Eur. Patent 662466, 1995. 73 W.J. Gamble, A.A. Naujokas, and B.R. De Bruin, US Patent 5 298 530, 1994. 74 D. Paszun and T. Spychaj, Ind. Eng. Chem. Res., 1997,36, 1373. 75 V. Matthews, Conference on Recycling of PET Containers, Maastricht, 1996. 76 H.L. Heiss, US Patent 3 109 824, 1963. 77 W.R. McElroy, US Patent 3 300417, 1967. 78 0. Kinoshita, US Patent 3 632 530, 1972.

58

Chapter 2

79 F.F. Frulla and R.R. Saylgh, US Patent 3 738 946, 1973. 80 B. Tucker and H. Ulrich, US Patent 3 983 987, 1976. 81 H. Ulrich, A. Odinak, B. Tucker, and A.A.R. Sayigh, Pofym.Eng. Sci., 1978,18(1l), 844. 82 H. Ulrich, Adv. Urethane Sci. Technol., 1978,5,49. 83 J. Braslaw and P. Pai, US Patent 4 159 972, 1979. 84 T. Miinzmay, H. Nefzger, W. Rasshofer, and W. Meckel, US Patent 5 508 3 12,1996. 85 L.C. Pizzini and J.T. Patton, Jr., US Patent 3 441 616, 1969. 86 B. Dawson, S. Hopkins, and P.R. Sewell, J . Appf. Polym. Sci., 1970,14, 35. 87 L.R. Mahoney, S.A. Weiner, and F.C. Ferris, Environ. Sci. Technol., 1974,8, 135. 88 G.A. Campbell and W.C. Meluch, Environ. Sci. Technol., 1976, 10, 182. 89 D.F. Lohr, Jr. and E.L. Kay, US Patent 4035 314, 1977. 90 R. C. Poller, J . Chem. Tech. Biotechnol., 1980,30, 152. 91 J.L. Gerlock, J. Braslaw, L.R. Mahoney, and F.C. Ferris, J . Appl. Pofym.Sci., 1980, 18, 541. 92 M.B. Sheratte, US Patent 4 110 266, 1978. 93 H. Lentz and W. Mormann, Makromol. Chem., Macromof. Symp., 1992,57,305. 94 J. Braslaw and J.L. Gerlock, Ind. Eng. Chem. Process Des. Dev., 1984,23, 552. 95 J.L. Gerlock, J. Braslaw, and J. Albright, US Patent 4 316 992, 1982. 96 J.L. Gerlock, J. Braslaw, and J. Albright, US Patent 4 317 939, 1982. 97 0. Kondo, T. Hashimoto, and H. Hasegawa, US Patent 4 014 809, 1977. 98 H.R. van der Wal, US Patent 5 274004, 1993. 99 K.V. Datye, Indian J . Fibre Textile Res., 1991,16(l), 46. 100 N. Chaupart, G. Serpe, and J. Verdu, Polymer, 1998,39(&7), 1375. 101 R.J. McKinney, US Patent 5 302 756, 1994. McKinney, US Patent 5 395 974,1995. 102 R.J. 103 G.A. Kalfas, Polym. React. Eng., 1998,6(1), 41. 104 P. Miiller and R. Riess, Makromol. Chem., Macromof. Symp., 1992,57, 175. 105 L.-C. Hu, A. Oku, and E. Yamada, Polymer, 1998,39(16), 3841.

CHAPTER 3

GasiJication and Partial Oxidation


1 Introduction
Gasification can be considered to be a partial oxidation process of carbonaceous materials leading predominantly to a mixture of carbon monoxide and hydrogen, known as synthesis gas or syngas due to its application in a variety of chemical syntheses. Gasification was initially developed for coal conversion, but it has been further applied to the processing of heavy petroleum fractions and natural gas. In the last two decades, gasification has also been investigated as a method of obtaining valuable chemicals from both biomass-derived products and organic solid residues. Currently, gasification can be considered an efficient treatment for the degradation and conversion of polymeric wastes. One of its major advantages lies in the fact that it is not necessary to separate the different polymers present in the plastic wastes. Moreover, in many cases plastic wastes are gasified while mixed with other components of the solid waste stream. However, the economics of a gasification process largely depend on the value and possible applications of the synthesis gas, either as an energy source by combustion or for the synthesis of various chemicals (methanol, ammonia, hydrocarbons, acetic acid, etc.). Only in the last case is gasification really a feedstock recycling method for plastic materials. In addition to gasification, other oxidative treatments of plastic and rubber wastes, excluding total combustion, are described in this chapter. These methods, although relatively unknown, may be of great interest in the future for the chemical degradation of polymeric wastes.

2 Gasification of Carbonaceous Materials and Uses of the Syngas


Gasification involves the reaction of the raw carbonaceous material with oxygen, air, steam or steam/oxygen and steam/air mixtures. It may take place over a wide range of temperatures (700-1 600 C) and pressures (10-90 atm).
59

60

Chapter 3

When oxygen or air are used as gasification agents, their content in the reaction medium must be kept low in order to avoid complete oxidation into carbon dioxide and water. The process can be promoted by metal catalysts, usually added to the raw material in aqueous solutions. The basic reactions which take place during the gasification of a carbonaceous material in the presence of oxygen and/or steam are shown in Scheme 3.1. In many cases, prior to the reactions involving oxygen or steam, the raw material undergoes a thermal decomposition that can be schematically expressed through reaction 1, yielding carbon and hydrogen. Conversion through partial or total oxidation (reactions 2-6) occur through exothermic transformations. Because gasification takes place in an oxygen-deficient atmosphere, partial oxidation reactions are predominant. On the contrary, those reactions involving the participation of steam are endothermic (reactions 7-9), and are mainly responsible for the formation of hydrogen. Reaction 8 is usually known as the water-shift reaction because it allows control of the H&O ratio.
Raw material decomposition

C H O X C + ~ H ,
X Y

Reactions with oxygen


1 c+0, o c o 2 1 co+0, O C O ,

I H, +- 0, a H,O 2

Reactions with water

C+ H,OaCO+H, CO+ H,O o CO, + H ,

(7)
(8)

(9)

Reactions with COI

Methanation reactions

Scheme 3.1 Basic reactions during the gasification of a carbonaceous material.

Gasification and Partial Oxidation

61

Other important endothermic transformations are those involving C 0 2 (reactions 10 and 11). The Boudouard reaction occurs between carbon and carbon dioxide to increase the yield of carbon monoxide (reaction 10). Finally, some methanation reactions by hydrogenation of carbon oxides also take place (reactions 12 and 13), which may lead to a significant decrease in the H2 concentration of the final synthesis gas. The overall energy requirement of gasification can be balanced by a suitable combination of exothermic and endothermic reactions, mainly through the control of the 02/H20 ratio in the reaction medium. Most of these processes are equilibrium reactions, which are established in the gasifier when working at temperatures between 1300 and 1500 "C. Below 900 "C, equilibrium can be approached only when operating in the presence of catalysts or with long residence times. As a result, the final gas is a mixture of H2, CO, C02, H 2 0 and minor amounts of other components such as CH4. The exact composition of the synthesis gas depends on a number of factors, the main ones being the reaction temperature and the 0 2 / H 2 0 ratio. According to the thermodynamic equilibria, an increase in the temperature causes a reduction in the H2 and C 0 2concentrations, whereas the formation of CO is enhanced.' The CO/H2 molar ratio can be controlled through the type and concentration of the gasification agent and can be balanced through the water-shift reaction. In addition to synthesis gas, a solid residue is also obtained, which contains all the inorganic elements that have not been gasified. Typically, the solid residue of the gasification processes is highly stable against leaching, which facilitates its further disposal.2 Nevertheless, depending on the reaction conditions, the gasification residue may vary from one similar to that obtained in pyrolysis treatments to one similar to an incineration residue. Depending on the carbonaceous solid used as raw material, various pollutants may be present in the gaseous products. Thus, H2S and carbonyl sulfide are found in significant amounts when starting from coal, petroleum residues and vulcanized rubber wastes. Likewise, in the treatment of plastic wastes the most noxious components are C1 and N compounds formed from the PVC, polyurethanes and polyamides present in the raw waste stream. Chlorine is usually released in the gaseous stream as HC1, whereas the nitrogen can be found as N2 or NH3. Accordingly, a crucial aspect of gasification is the cleaning of the gases in order to remove all these undesired components, which is usually achieved by several processes: absorption with or without subsequent reaction, catalytic conversion, adsorption, separation by membranes, etc. The energy content of the gaseous stream leaving the gasification reactor is usually utilized in the generation of steam, which can be used as the gasification agent or as a heat source in the reactor. After cooling and cleaning, the gas from a gasification reactor can be employed in a number of applications:
0

Energy recovery. Synthesis gas can be used as fuel gas. In this case the presence of large amounts of methane is advantageous, which can be achieved by varying the reaction conditions in the gasifier to promote methanation reactions by hydrogenation of CO and C02.

62
0
0

Chapter 3

0 0

Reducing gas for ferrous and non-ferrous ores. Synthesis of methanol by catalytic hydrogenation of CO and C02. A higher H2/CO ratio is necessary for this process, which can be obtained by promoting the water-shift reaction. Ammonia synthesis, which also requires an increase in the H2 concentration through the water-shift reaction. Synthesis of hydrocarbons by the catalytic Fisher-Tropsch process. 0 x 0 synthesis. Hydroformilation of or-olefins with CO and H2 is usually used for the synthesis of terminal oxygenated compounds (alcohols, aldehydes and carboxylic acids).

The large number of applications of the synthesis gas produced in gasification is one of the reasons for the interest in this technology and its potential for the conversion of carbonaceous materials. Another important factor is its flexibility with respect to the type, characteristics and contamination level of the material to be processed. However, one of the main drawbacks of this method is that, in order to ensure the economic profitability of gasification, large plants with capacities of about 400 000-500 000 tonnes/year must be in ~ p e r a t i o n . ~ Brief descriptions of commercial gasification processes can be found in the gasification with steam (Exxon process), gasification with oxygen or air (Koppers-Totzek and Prenflo processes), and gasification with steam/oxygen or air mixtures (Shell, Winkler and Texaco processes).

3 Gasification of Plastic and Rubber Wastes


Most of the processes so far proposed for the gasification of polymeric wastes have been directly derived from earlier processes developed for the conversion of coal, natural gas and heavy petroleum fractions. However, certain details must be taken into account when processing plastic and rubber wastes in the gasification units, for instance the heterogeneity of the starting material, the problem of feeding the highly viscous melted plastics, and the possible formation of corrosive compounds, mainly HCl from PVC. Bockhorn et al.9 have investigated the partial oxidation of different polyolefins (polyethylene and both atactic and isotactic polypropylene) with oxygen in fuel-rich flames. A scheme of the experimental apparatus used is shown in Figure 3.1. The polyolefinic plastics are first liquefied in a melting vessel and are then pumped by a reciprocating pump into the combustion chamber, where they are atomized with a gas mixture consisting of hydrogen and carbon monoxide (1:l mole ratio) and contacted with oxygen. Figure 3.2 illustrates the plain-jet atomizer used to obtain good mixing of the reactants and to assist in the dispersion of the plastic droplets and their contact with the oxidant. The gaseous products of the partial oxidation of the polyolefins were mainly H2 and CO (global concentrations between 50 and 60%), although significant amounts of methane, ethylene, acetylene, carbon dioxide and water were also detected. One of the most important variables determining the gas

t o Gas chromatograph R

Melting Vessel

Figure 3.1 Experimental set-up for the partial oxidation of polyolejins9: pressure reduction valve ( I ) , needle valve ( 2 ) , rotameters ( 3 ) , slide valve ( 4 ) , reciprocating pump ( 5 ) , pulsation absorber (6). (Reprinted from J. Anal. Appl. Pyrol., 8, H. Bockhorn, M. Burckschat and H. Deusser, page 427. 0 1985, with permission from m Elsevier Science) w

64

Chapter 3

Liquid Pdyolefins

Figure 3.2 Plain-jet atomizer f o r the dispersion and partial oxidation of polyolejiin~.~ (Reprinted from J. Anal. Appl. Pyrol., 8, H. Bockhorn, M . Burckschat and H. Deusser, page 427. 0 1985, with permission from Elsevier Science)

composition was the mass flow ratio of polyolefin/oxygen. As this ratio was varied over the range 0.61-1.02 in the conversion of atactic polypropylene, the methane and ethylene concentrations decreased from 9.6 to 4.1YOand from 5.1 to 0.9%, respectively. Lower conversions of polyolefins were obtained compared to the treatment of C14-CI7 n-alkanes, because the polymers are less finely atomized and more slowly vaporized. These results indicate that the conversion rate of plastics with this system depends on the ease of atomization and vaporization of these fuels. The problems associated with the gasification of C1-containing plastic wastes has been addressed in a recent work." The author proposed a two-step process for the conversion of PVC wastes: (i) PVC dehydrochlorination by a thermal treatment, and (ii) gasification of the residues so obtained with a mixture of steam and oxygen. Thermal analysis of the raw PVC confirmed that PVC dehydrochlorination and pyrolytic degradation occur at different temperature levels, which makes it possible to isolate each process (see Chapter 4). Thus, treating PVC at 350C in an inert atmosphere allows most of the C1 to be removed as HCl, with a concentration in the volatile products of over go%, which makes its chemical reutilization feasible. Regarding the second step, the author simulated the equilibrium corresponding to the gasification of thermally treated PVC residues with C1 contents in the range 0.01-0.5 wt%, assuming it has a structure similar to polyacetylene. These calculations showed that the remaining C1 is removed during gasification mainly as HC1, although dioxins may also be formed. Thermodynamic calculations showed that when the 0.5 wt% C1-containing residue is gasified at temperatures above

Gasification and Partial Oxidation

65

1000 "C, the concentration of 2,3,7,8-tetrachlorodioxinin the obtained gas may reach levels of several mg/m3. Moreover, within the range of investigated temperatures, the estimated concentration of dangerous dioxins increased with the temperature. These results show that during the processing of PVCcontaining wastes by gasification, it is desirable to reduce the C1 content as far as possible before the wastes enter the gasifier in order to ensure that dioxins are not formed. Kowallik et a1.l' have recently patented a process for the simultaneous gasification of both plastic wastes and heavy oils, in which these materials are reacted with oxygen in the presence of steam as a moderating agent at temperatures in the range 1300-1500 "C and pressures between 40 and 60 atm. Plastic and oil wastes are fed to the gasifier as two independent streams. The gases leaving the reactor are passed through a heat recovery unit and subsequently the unreacted carbon and ash are removed by water washing. The relative proportion of oil and plastics can be widely varied up to a maximum plastic content of around 50 wt%. The authors also claim that this process can tolerate C1-containing plastics to a certain extent, the resulting HC1 being disposed of by neutralization. A combination of thermal and gasification steps takes place in the process proposed by Menges3 for the conversion of plastic wastes (Figure 3.3). In this case the plastic waste stream first undergoes a degradative extrusion at 400 "C, which causes a partial depolymerization and a significant reduction of the melt viscosity. The material is then mixed with oxygen and steam and gasified at 1400 "C and 50 bar. The synthesis gas obtained from the gasifier is cleaned in a scrubber, and could possibly then be used in hydrocarbon synthesis. A somewhat different approach to the gasification of plastic wastes is described in the patent of Khan et a1.,12 assigned to Texaco. In this case, the plastic material is first granulated and mixed with liquid water to produce a plastic sludge with a solids content between 60 and 80 wt%. This mixture is hydrothermally treated at temperatures in the range 230350"C, which causes a certain degradation and the formation of gases containing CO, CO2, H*S, NH3 and light hydrocarbons. After the separa-

oxygen

water vapour waste h e a t boiler soot, slag, condensate

clean synthesis g a s

I
water

Figure 3.3 Simplijiedflow diagram of a process for the gasification of plastic

waste^.^

66

Chapter 3

tion of these gases, the resulting sludge is mixed with a carbonaceous material and additional water to obtain a pumpable aqueous slurry with a solids content up to 60%. The carbonaceous solid material can be particulate carbon, coal, coke from coal, petroleum coke, oil shale, tar sands, asphalt, pitch, etc. Therefore, this process allows the simultaneous gasification of plastic wastes and a variety of carbonaceous solids. Gasification of this mixture takes place with an 02-containing gas at temperatures between 1300 and 1540 "C, yielding a gaseous product containing CO, H2, CO2, CH4 and H2S. When treating plastics that contain halides, such as PVC or PTFE, these are released as HCl and HF, respectively. Plastics often also include Br-containing fire retardants, which lead to the formation of HBr. These undesirable hydrogen halides can be removed from the gaseous stream by absorption on water containing ammonia. All the inorganic elements present in the raw plastic wastes and carbonaceous materials are recovered from the gasifier as a non-toxic, non-leachable slag, which can be considered as a by-product with applications in road beds and building blocks. According to the authors, this process can be applied to a wide variety of polymeric wastes, without limitations in regard to the presence of high amounts of fillers, pigments and reinforcing agents. Closely related to the previous process is the plastic waste gasification facility being set up at Rotterdam with a capacity of 150 tonnes day-' of plastic wastes.I3 The feed of this plant will be a mixture of polyethylene, polypropylene and polystyrene with minor amounts of other polymers (2.4 wt% of PVC) and a significant proportion of cellulose. The plan is that an injection of ammonia into the gasifier will neutralize the chlorides entering with the plastic wastes, which will lead to the formation of ammonium chloride salt as a by-product. A production of 350 000 m3 day- of synthesis gas is estimated, which will be used in chemical synthesis. In the case of the gasification of rubber wastes, Ogasawara et a l l 4 have studied the decomposition of used automotive tyres by treatment with steam at different temperatures. The main components detected in the effluent gases were H2, CO and CH4. As shown in Figure 3.4(a), the production rates of both hydrogen and carbon monoxide increase exponentially with the temperature. The H 2 / C 0 ratio is about 1.5 whereas CH4 is produced mainly at long reaction times. An oil fraction is also produced, containing mostly aliphatic hydrocarbons and alkylbenzenes. The solid residue obtained by Ogasawara et a1.14 in the gasification of used tyres with steam has a high surface area, hence it can be considered a sort of activated carbon. Treatment at 900C for 1 h leads to a carbon residue with a surface area of 1260 m2 g-' with a yield of 9%. Benzene adsorption measurements carried out on this sample showed that it can be considered a good activated carbon. The authors found a linear relationship between the carbon residue yield and its surface area, independently of the treatment conditions, as can be seen in Figure 3.4(b). The low yields of carbon residue obtained at high temperatures indicate that gasification affects both the rubber and the carbon black present in the raw tyre.

Gasification and Partial Oxidation


VI

67
B)

900

1000

1100

1200

20

40

Reoctlon t m r a t u r e IKI

\ileld of corbon residue I:

Figure 3.4

GasiJication of used tyres with steam:I4 (A) rate formation of H2 and CO versus temperature, (B) relationship between the surface area and the yield of the carbon residue. (Reprinted with permission from S. Ogasawa, M. Kuroda and N. Wakao, Ind. Eng. Chem. Res., 26,2556. 0 1987 ACS)

4 Gasification of Mixed Solid Wastes


One of the major advantages of gasification for the conversion of wastes is that it requires little pretreatment or sorting of the starting materials. Consequently, in many gasification processes plastic and rubber wastes are converted simultaneously with other solid wastes. This section describes some of the processes currently used for the gasification of mixed solid wastes. Lutge et ~ 1 . have ' ~ described the Krupp Uhde PreCon process, based on the high temperature Winkler gasification, for the conversion of solid wastes such as municipal solid waste, sewage sludge, auto shredder residue, plastic wastes, etc. The raw material is gasified at pressures up to 30 atm and temperatures between 800 and 1100 "C in a fluidized bed with oxygen or air, being fed into the gasifier by a screw feeder (Figure 3.5). The use of either air or oxygen depends on the required quality of the produced gas. The products are synthesis gas and a residue containing unreacted carbon and ash, which is removed by a screw located in the bottom of the gasifier. Detailed analysis of the products obtained in demonstration plants has shown that no dioxins or furans are present, thought to be due to the presence of a reducing atmosphere in the gasifier. Compared with conventional incineration processes, this method results in smaller gas volumes, which decreases considerably the costs of gas treatment. TPS Termiska Processer AB has been working since the mid-1980s on the development of an atmospheric gasification process consisting of two steps. l 6 Gasification of the raw material takes place in a circulating fluidized bed operating under atmospheric pressure and temperatures in the range 850900 "C with air as gasification and fluidizing agent. The gas leaving the gasifier is fed into a second fluidized bed reactor where the heaviest hydrocarbons are cracked into lighter components by contact with a dolomitic catalyst. The HC1

68
Crude Gas Cooler

Chapter 3

Charge

Fluidizec
Feed Sc

(0, / Steam or Air) (0, / Steam or Air)


Ash Discharge Sys

Gasification I agent

Gasification agent

ollection Bin

Lock Hopoer

Figure 3.5

Gasifier of the high temperature Winkler process f o r the conversion of solid wastes. l 5 (From C . Lutge, M. Deutsch, P. Wischnewski, H.-P. Schiffer and D. Hoey, Proceedings of the International Conference on Incineration and Thermal Treatment Technologies, Salt Lake City, Utah. 0 1998 University of California, Irvine Office of Environment, Health and Safety)

formed is absorbed by the calcined dolomite to form CaC12, which is further removed in a downstream filter. The Thermoselect process is another interesting alternative developed for the gasification of wastes.' Figure 3.6 shows the different steps of this process. The waste is first compressed and compacted in order to increase its bulk density, remove entrained air and homogenize the material. Subsequently, the waste is fed into a pyrolysis channel where it is degasified and thermally decomposed as the temperature is progressively increased up to 600 "C with a residence time of about 2 h. The pyrolysis products enter the gasifier where they are reacted with oxygen and steam. The organic components are transformed into synthesis gas which leaves the reactor at 1200 "C, while the inorganic portion is converted into liquid melt phases in the bottom of the gasifier at 2000 "C,flowing slowly into a homogenization reactor. The residence time of the gas phase in the gasifier is about 2-4 s, whereas the liquid molten phase remains for about 5 min. The synthesis gas produced is subjected to shock cooling by contact with water to prevent the possible formation of dioxins, furans and other undesired organic compounds, as well as to absorb most of the HCI present. Further purification treatments are applied to the synthesis gas to remove fine particles, H2S, water vapour, etc. Other gasification processes developed for the conversion of different types of wastes have been described by Whiting,17 whereas a summary of the most

Compression

- Degasing

- Gasification

Syngas for use

Figure 3.6

Simplijied flow diagram of the Thermoselect process f o r the gasijication of wastes.

'

relevant commercial processes existing for the gasification of plastic wastes can be found in the work of Hofmann and G e b a ~ e r . ~

5 Other Plastic and Rubber Partial Oxidation Processes


Partial and controlled oxidation of polyolefinic plastics is used to promote polymer crosslinking in order to enhance their physical properties and improve their processability. Thus, a number of have been published on the treatment of polyethylene and polypropylene with different peroxides, which usually results in an increase in the polymer melting temperature and a narrowing of the molecular weight distribution. On the other hand, in the thermal degradation of polyolefins, it is well known that the presence of low amounts of oxygen accelerates polymer degradation, whereas at higher oxygen concentrations complete oxidation into carbon dioxide and water is predominan t. In spite of the above mentioned background on polymer oxidation, only a few works have been published on the subject of obtaining commercially valuable products by partial oxidation of plastic and rubber wastes. Several of these works are described below. Watabe et af.21have developed a process for the oxidative decomposition of vulcanized rubber by treatment with an organic hydroperoxide under mild reaction conditions. It can be applied to both natural and synthetic rubbers, and a wide range of organic peroxides can be used (tert-butylhydroperoxide, cumene hydroperoxide, diisopropylbenzene hydroperoxide, etc.). The hydroperoxide is contacted with the vulcanized rubber in a proportion of between 1 and 5 wt% in the presence of halogenated hydrocarbons as solvents and Cu or Co salts as oxidation catalysts. The reaction proceeds at temperatures around 40 "C and is accelerated by a supply of air or oxygen. Under these conditions, total decomposition of the rubber can be achieved after several hours of reaction. A number of possible applications are proposed by the authors for the resulting solution: process oil, softener, compounding ingredient for rubber, modifier for asphalt, etc.

70

Chapter 3

Gebauer and Hofmann22 have suggested a method for promoting plastic decomposition by thermooxidation. Treatment of polyolefins with oxygen at relatively low temperature (130-1 80 "C) causes rapid degradation of the polymers to yield wax oxidates. One of the main problems with this method is the dispersion and contact of the oxygen with the highly viscous melted polymer. According to the authors, the wax oxidates obtained present interesting properties for a variety of applications. A completely different approach to plastic oxidative depolymerization is the . ~ ~ of polymer degradation method recently proposed by Lee et ~ 1 It consists through its selective partial oxidation with oxygen in a supercritical water mixture or in a water mixture near supercritical conditions, which leads to high yields of the starting monomer. According to the authors, this process can be successfully applied to both condensation and addition polymers, examples being provided for the oxidative degradation of PVC, isotactic polypropylene and polystyrene. The polymer is rapidly brought to conditions above the water critical temperature (374 "C) and pressure (218 atm) by contact with supercritical water, after which oxygen is injected into the reactor. Under these conditions, the oxygen is completely miscible with the water, creating a single fluid phase, whereas the polymer swells and partially dissolves, which results in the formation of a pseudo-homogeneous mixture. Two types of reaction take place in this system: polymer decomposition by chain scission and total oxidation. In order to reduce the latter, relatively small quantities of oxygen are introduced into the reactor. The degradation is very fast, with most of the polymer being completely converted within a residence time of 3 min, while the presence of a catalyst is not required. Typical products of polymer oxidation under supercritical water conditions include the corresponding monomers, dimers, trimers, and other oligomers having up to 8 repeat groups, as well as carbon dioxide, water and trace amounts of light hydrocarbons. If the polymer contains organic additives, they are also subjected to oxidative degradation, whereas inorganic additives and impurities present in the starting polymer are not generally soluble in the supercritical fluid mixture, and are easily separated from the reaction product as a precipitate. In the oxidative depolymerization of isotactic polypropylene, the main products obtained are propylene, methane, ethylene and acetic acid. Propylene yields of up to 65 wt% are produced. Likewise, polystyrene oxidative degradation under supercritical water conditions leads to methane, ethylene, propylene, isobutene, propane, butane, benzene, toluene, ethylbenzene and styrene. A more complex degradation takes place when this process is applied to PVC. The authors propose that PVC depolymerization under supercritical water conditions proceeds in accordance with a mechanism consisting of four different pathways: (i) dehydrochlorination and partial oxidation, (ii) dehydrochlorination and chain scission, (iii) dehydrochlorination and total oxidation, and (iv) hydrochlorination. In the reaction products, high yields of vinyl chloride, 1,1-dichloroethane and 1,2-dichloroethane are detected, especially at short reaction times, whereas longer times favour total oxidation products.

Gasijication and Partial Oxidation

71

Other products found include chloromethane, chloroethane and 1,2-dichloropropane. In a further patent,24 this process has been extended to the oxidative decoupling of scrap rubber under water supercritical conditions. Both natural and synthetic, vulcanized and non-vulcanized, rubbers have been processed to yield a mixture of lower alkanes, alkenes, dienes, aromatics, alcohols, carboxylic acids, aldehydes and ketones. In addition, small quantities of rubber fragments, carbon dioxide, water, sulfur and nitrous oxides, and halide acids can also be formed, depending on the raw rubber. The reaction takes place above the critical water temperature and pressure in the presence of oxygen. Under these conditions, the rubber polymers are very rapidly broken down through chain scission (breakage of carbonxarbon bonds), and devulcanization when starting from a vulcanized rubber (breakage of sulfur-sulfur and carbon-sulfur bonds), as well as total oxidation into carbon dioxide and water. It is proposed that the resulting products can be used as a fuel, with or without subsequent upgrading to obtain a fuel of higher quality, or to produce various chemicals by separation of its components.

6 Summary
Gasification of plastic wastes by treatment with oxygen, air, steam or mixtures of these agents is an effective degradation method that yields synthesis gas and a solid residue as the main products. Gasification can be viewed as a feedstock recycling process for plastic wastes provided that the synthesis gas produced is used in the preparation of methanol, ammonia, hydrocarbons, oxygenated compounds, etc., instead of being used as a fuel gas. The major advantage of gasification is the possibility of treating complex mixtures without any previous separation step. Thus, in many cases plastics are gasified while mixed with other solid wastes. The main drawbacks are related to the downstream cleaning operations which are required to produce a syngas with quality and composition suitable for chemical synthesis. Moreover, the economic viability of most chemical processes starting from synthesis gas strongly depend on their application in large-scale plants located close to the facilities where the wastes are converted into the gas. Other possible oxidative treatments for the feedstock recycling of plastic and rubber wastes include partial oxidation with organic peroxides and decomposition by reaction with oxygen by thermooxidation or under supercritical water conditions. However, these latter alternatives have so far not been widely investigated.

7 References
1 B. Calaminus and R. Stahlberg, Proceedings of the International Conference on

Incineration and Thermal Treatment Technologies, Salt Lake City, Utah, 1998, p. 127. 2 A.G. Buekens and J.G. Schoeters, Conserv. Recycling, 1986,9(3), 253.

72

Chapter 3

3 U. Hofmann and M. Gebauer, Kunststofe Germun Plrst., 1993, 83,4, 4 Hydrocarbon Processing, 1984,63(4), 90. 5 J. Anwer, I.N. Banchik, G.W. Bode, W. Lemberg, and K.K. Sud, US Patent 4 0 17 272, 1977. 6 C.L. Reed and C.J. Kuhre, Hydrocarbon Processing, 1979, 58(9), 191. 7 R.A. Cardello, R.A. Reitz, and F.C.R.M. Smith, Proceedings of the 1 lth World Petroleum Congress, London, 1983. 8 K.A. Theis and E. Nitschke, Hydrocarbon Processing, 1982,61(9), 233. 9 H. Bockhorn, M. Burckschat, and H. Deusser, J . Anal. Appl. Pyrol., 1985,8,427. 10 K. German, 12th International Congress of Chemical and Process Engineering, Praga, 1996. 11 W. Kowallik, H.J. Maaz, and W. Soyez, WO Patent 11465, 1994. 12 M.R. Khan, C.C. Albert, and S.J. De Canio, WO Patent 9903, 1995. 13 R.C. Weissman, in Chemical Aspects of Plastics Recycling, ed. W. Hoyle and D.R. Karsa, The Royal Society of Chemistry, Cambridge, 1997, p. 199. 14 S. Ogasawara, M. Kuroda, and N. Wakao, Ind. Eng. Chem. Res., 1987,26,2556. 15 C. Lutge, M. Deutsch, R. Wischnewski, H.-P. Schiffer, and D.L. Hoey, Proceedings of the International Conference on Incineration and Thermal Treatment Technologies, Salt Lake City, Utah, 1998, p. 135. 16 M. Morris and L. Waldheim, Proceedings of the International Conference on Incineration and Thermal Treatment Technologies, Salt Lake City, Utah, 1998, p. 141. 17 K.J. Whiting, Waste Management, 1997,2, 150. 18 H.J. Riddle, SOC.Plast. Eng. Antec., 1980,38, 183. 19 P. Hudec and L. Obdrzalek, Angew. Makromol. Chem., 1980,89,41. 20 J. De Boer, and A.J. Pennings, Makromol. Chem. Rapid. Commun., 1981,2(12), 749. 21 Y. Watabe, H. Takeichi, and K. Irako, US Patent 4426459, 1984. 22 M. Gebauer and U. Hofmann, Proceedings of Recycle 93, Davos, Switzerland, 1993. 23 S. Lee, M.A. Gencer, K.L. Fullerton, and F.O. Azzam, US Patent 5 386055, 1995. 24 S. Lee, F.O. Azzam, and B.S. Kocher, US Patent 5516952, 1996.

CHAPTER 4

Thermal Processes
1 Introduction
This chapter describes the different methods and processes developed for breaking down polymeric materials simply by treatment at high temperature in an inert atmosphere. Studies on the thermal degradation of plastics have been reported for many years, a huge number of references on this topic being found in the literature. Moreover, several books have been published on the thermal degradation of plastic and rubber materials.'-'' Most of the early studies on polymer thermal degradation were aimed at determining the polymer thermal stability rather than at developing feasible alternatives for the conversion of polymeric wastes. It was in the 1970s that the thermal degradation of plastics and rubber began to be considered as an interesting alternative for the feedstock recycling of such wastes, Polymer decomposition by treatment at high temperatures is also used as an analytical technique for polymer identification. , 1 2 Thermal decomposition of polymers can be considered as a depolymerization process in only a few cases. Thus, polystyrene and polymethylmethacrylate are examples of polymers that can be thermally degraded with the formation of high yields of the corresponding monomer. However, for most polymers thermal decomposition leads to a complex mixture of products, containing low monomer concentrations. The type and distribution of products derived from the thermal degradation of each polymer depends on a number of factors: the polymer itself, the reaction conditions, the type and operation mode of the reactor, etc. Of the reaction variables, it is obvious that temperature is the most significant, because it influences both the polymer conversion and the product distribution. In general terms, up to four product fractions can be recovered from the thermal decomposition of plastics and rubber materials, depending on their physical state at room temperature: gases, oils, solid waxes and a solid residue. As the temperature is increased, the fraction of gases also increases and the solid residue appears as a solid char due to the enhancement of hydrocarbon coking reactions. Therefore, a variety of products and applications can be envisaged from the thermal decomposition of polymeric materials: fuel gases, olefinic gases useful in chemical synthesis, naphtha and middle distillates, oil

''

73

14

Chapter 4

fractions, long-chain paraffins and olefins, coke, etc. These products can be directly used as fuels or as a source of chemicals or alternatively they can be further processed and upgraded in refineries to yield higher quality fuels. Some confusion is found in the literature over the terms used to describe the thermal treatment of polymers: depolymerization, cracking, thermolysis, pyrolysis, etc. In this book, we shall use the term pyrolysis to refer to the thermal decomposition of polymeric materials at high temperatures (above 600 "C) whereas, when the degradation takes place at lower temperatures, we shall refer to it mainly as thermal cracking. However, in some cases it is difficult to assign a process to one of these categories, as is the case in thermogravimetric analysis where the temperature is continuously varied. Thermal processes are mainly used for the feedstock recycling of addition polymers whereas, as stated in Chapter 2, condensation polymers are preferably depolymerized by reaction with certain chemical agents. The present chapter will deal with the thermal decomposition of polyethylene, polypropylene, polystyrene and polyvinyl chloride, which are the main components of the plastic waste stream (see Chapter 1). Nevertheless, the thermal degradation of some condensation polymers will also be mentioned, because they can appear mixed with polyolefins and other addition polymers in the plastic waste stream. Both the thermal decomposition of individual plastics and of plastic mixtures will be discussed. Likewise, the thermal coprocessing of plastic wastes with other materials (e.g. coal and biomass) will be considered in this chapter. Finally, the thermal degradation of rubber wastes will also be reviewed because in recent years much research effort has been devoted to the recovery of valuable products by the pyrolysis of used tyres.

2 Mechanism of the Thermal Degradation of Addition Polymers


Thermal degradation of plastics and rubber proceeds through a radical mechanism, which may involve three different decomposition pathways: (i) Random scission at any point in the polymer backbone leading to the formation of smaller polymeric fragments as primary products, which in turn may be subjected to additional random cracking reactions. (ii) End-chain scission, where a small molecule and a long-chain polymeric fragment are formed. If the small molecule released is the starting monomer, the thermal degradation process can be considered as an actual depolymerization or unzipping process. (iii) Abstraction of functional substituents to form small molecules. In this case, the polymer chain may retain its length or the release of the small molecule may be accompanied by cleavage of the polymeric chain. In many cases several of these pathways occur simultaneously. Thus, polyethylene and polypropylene are thermally degraded by both random and end-

Thermal Processes

75

chain scissions. In the case of PVC, however, the predominant mechanism of the first step is the removal of HCl according to pathway (iii), followed by the decomposition of the remaining diene backbone at a higher temperature, mainly via pathway (i). During the thermal degradation of many polymers other reactions may take place at the same time as the cracking reactions, e.g. isomerization, cyclization, aromatization, recombination of species, etc. Thus, an increase in the degree of branching of the polymeric chains is usually observed as they are reduced in length by thermal decomposition. According to the pathways described above, the thermal decomposition of polymers often involves the formation of volatile species within a highly viscous polymeric matrix. Transport of these species through the molten polymer mass towards the vapour phase is not a straightforward process, and so the occurrence of mass transfer limitations can be expected. Various authors have observed that the rate of polymer thermal degradation depends on factors such as the surface area and thickness of the polymer sample, showing that the polymer decomposition is controlled by the diffusion and/or the vaporization of the volatile On the other hand, polymer thermal decomposition is an endothermic process. At least the dissociation energy of the C-C bond in the chain must be supplied to break down the polymer. Moreover, this is the main factor determining the polymer stability. Thus, as shown in Figure 4.1, a direct relationship between the dissociation energy and the decomposition temperature has been found for different polymers.'' Because molten polymers are highly viscous fluids with low thermal conductivity, heat transfer limitations may also be present, leading to significant temperature gradients. However,

550
L

-~

I' 1

T,("C) 500 450

PTFE

PE

400

350 -

300

PMS

/
1

250

50

60

70

80

90

Dissoc. energy (kcal/mol)

Figure 4.1 Relationship between the decomposition temperature and the dissociation energy o f the C-C chain bonds for diflerent polymers.

Initiation
. b
w , v

Random scission:

C H2 - CHX - C H2 - CHX ww C H2 - 6HX

End-chain scission:
+

~ M C H ~ - C H X - C H ~ - C Hd ~X 'WCHz-CHX

.
+

&H2-CHXbm$

t H 2 -CH2X

DeDrODaeatiOn
'WCH2-CHX

wCHZ-CHX-CHZ-CHX

.
+
CH2ZCHX

Hvdroeen chain transfer

Intermolecular:

Intramolecular:

P-cleavape

Formation of branches

Termination

Bimol ecular coupling:

Disproportionation:

Scheme 4.1

General mechanism for the thermal degradation of addition polymers.

Thermal Processes

77

this effect is not taken into consideration in most of the laboratory studies due to the use of small polymer samples, but it is a crucial aspect in the design of pilot plants or industrial reactors for the thermal conversion of plastic wastes. Scheme 4.1 shows a general mechanism proposed by Stivala et af. for the thermal degradation of addition polymers with the following steps: Initiation, involving the scission of the first bonds in the chain yielding two radicals, which may occur at random or end-chain positions. Depropagation, consisting of the release of olefinic monomeric fragments from primary radicals. Hydrogen chain transfer reactions, which may occur as intermolecular or intramolecular processes, as depicted in Scheme 4.1. This hydrogen abstraction by radicals leads to the formation of olefinic species and polymeric fragments. Moreover, secondary radicals can also be formed from hydrogen abstraction through an intermolecular transfer reaction between a primary radical and a polymeric fragment. P-Cleavage of secondary radicals to yield an end-chain olefinic group and a primary radical. Formation of branches by the interaction between two secondary radicals or between a secondary and a primary radical. Termination, which takes place either in a bimolecular mode, involving the coupling of two primary radicals, or by disproportionation of the primary macroradicals. The significant role played by the defects and impurities present in the polymeric backbones in the initial stages of thermal degradation is widely accepted. Polymer chains often contain small side branches, double bonds, oxygenated groups, etc. which act as weak links favouring the formation of the primary radicals and, therefore, promoting the initiation step of the polymer thermal degradation.

3 Thermal Conversion of Individual Plastics


This section reviews the different aspects of the thermal conversion of those polymers which are the main components of the plastic waste stream: polyethylene, polypropylene, polystyrene, PVC and PET, although the thermal degradation of other polymers is also commented on. The discussion focuses on mechanistic and kinetic factors, as well as on the type of products derived from the thermal decomposition of each individual polymer. The thermal degradation of plastic mixtures, which reflects more accurately the phenomena taking place in the thermal conversion of plastic wastes, is analysed and discussed in the next section.

78

Chapter 4

Polyethylene
Polyethylene is the major polymer present in plastic wastes. Both low density and high density polyethylene are found in large quantities in plastic residues. HDPE is a highly linear polymer, whereas LDPE possesses a certain degree of branching. HDPE exhibits both a higher crystallinity and a higher crystalline melting point than LDPE, because the linear chains of the former can be more closely packed. Differences are also observed in the thermal behaviour and stability of these polyolefins. Figure 4.2 compares the thermogravimetric analysis (TGA) of HDPE and LDPE in a nitrogen atmosphere. It can be seen that LDPE degradation takes place at lower temperatures than in the case of HDPE, which is probably related to the higher degree of branching present in LDPE, providing a higher proportion of reactive tertiary carbons for the initiation step of degradation. In both cases, the polyolefins are completely volatilized at temperatures below 500 "C.Abou-Shaaban et aZ.16 have proposed three stages to describe the polyethylene TGA curves. The first step, accounting for only around 3% weight loss, corresponds to the volatilization of low

100

200

300

400

500

600

700

O0 * , ...
80

I-

.. .-.......

417
100

IC
500

L
600

200

300

400

700

Ternperature (OC)

Figure 4.2

TG analysis of HDPE and LDPE in a nitrogen atmosphere.

Thermal Processes

79

molecular weight species due to the scission of most of the side chains. The second step is related to the cracking of the polymer backbone, whereas the third step is associated with the decomposition of the remaining carbonaceous residue. Various authors have proposed that PE thermal degradation proceeds mainly by a random chain scission mechanism to form intermediate species (heavy waxes and tars), which are further cracked to produce the final products (gases, aromatics, long-chain paraffins and olefins, coke, etc.).17-19 In other cases, it is assumed that end-chain cleavage takes place simultaneously to yield some of the observed gaseous products. Most of the kinetic studies on PE thermal degradation have been based on TGA measurements, mainly using a power law model to describe the rate of weight loss. In a recent paper,*' these studies have been reviewed and a new model (the RCD model) has been developed to describe the low temperature thermal cracking of polyolefins. The RCD model takes into account the influence of factors such as molecular weight, extent of branching and fl-scission, and evaporation of species from the reaction medium. However, the most restrictive limitation of the models based on TGA measurements is that they are of no help in the description of those steps which are not accompanied by a weight loss, as happens in the initial stages of the random chain cleavage. Thus, according to Figure 4.2, polyethylene volatilization begins at temperatures close to 400C, although the degradation of the polyolefinic chains is started at lower temperatures. Figure 4.3 illustrates the changes in the average molecular weight of LDPE, determined by gel permeation chromatography (GPC), and the amount of volatile products formed when this polymer is treated at different temperatures in a stirred tank reactor under a nitrogen atmosphere.2' Although the production of gases is negligible up to

-0-M , -0- Volatile Production (wt

YO)

- 60
- 50 - 40 - 30 - 20 - 10
-====-----""

225

250

275

300

325

350

3 7 5 -

4hO

0 475

VC)
Figure 4.3
Changes in the molecular weight andproduction o f volatiles during the LDPE thermal degradation .2'

80

Chapter 4

400 "C, from 350 to 400 "C the polymer undergoes significant degradation, leading to a large decrease in the average molecular weight. These transformations are not detected in the TGA measurements, because they take place mainly through a random scission mechanism to form polymeric fragments, which are not volatilized at these temperatures. The occurrence of end-chain cracking reactions or the removal of short-chain branches are probably the reasons for the formation of the gaseous products observed, mainly C1-C5 hydrocarbons. In a recent work,22a continuous kinetic model has been proposed to describe the changes occurring in the molecular weight distribution during the thermal cracking of PE at 370410C. The experiments were carried out in a novel reactor based on the stirring of the molten PE by bubbles of flowing nitrogen at atmospheric pressure, which allows low temperature gradients to be achieved. Both the products remaining in the reactor and those contained in the effluent stream were analysed. Molecular weight distributions were determined by GPC, indicating that random scission and repolymerization leading to crosslinking occur simultaneously, causing a broadening of the MWD. The model assumes that gaseous products are formed by end-chain cracking and takes into account the mass transfer of vaporized products from the polymer melt into gaseous bubbles, although the experimental data used did not permit estimation of the kinetic coefficients corresponding to the random scission and repolymerization processes. The products obtained by PE thermal decomposition largely depend on the degradation temperature and the reactor type. Most of the studies appearing in the literature on PE thermal decomposition focus on pyrolysis of the polyolefin by treatment at high temperatures (usually above 600C); only a few papers deal with the thermal cracking of PE at lower temperatures. Darivakis et al.23conducted PE pyrolysis in an electrically heated sample holder under helium flow. Polymer weight loss was almost total at 700 "C, two product fractions being recovered: gases and condensables, the latter being formed by tars and higher molecular weight volatilizable products. At a heating rate of 1000 "C s- PE volatilization approached 100% in just 0.7-0.8 s. Kamin~ky* has ~ investigated the pyrolysis of PE using a sand fluidized bed reactor. Figure 4.4 shows the evolution of the main products obtained at temperatures in the range 650-8 10 "C. These products are light hydrocarbons, with a high proportion of olefins (ethylene, propylene, cyclopentadiene, etc.) and aromatics (benzene and toluene). Methane and free hydrogen are also detected in significant amounts. Major changes in the product distribution are observed when the temperature is increased: the propylene content decreases, ethylene passes through a maximum, and benzene increases to reach a maximum value of about 25 wt%. A pronounced maximum is also observed in the methane curve with a 20 wt% yield at around 760 "C. PE pyrolysis in a fluidized bed reactor was also investigated by Scott et al.25 The major products were gaseous hydrocarbons with yields of around 60% at temperatures of 730 and 790 "C,although significant amounts of condensates (around 30%) were also obtained. The gas fraction was rich in olefins, especially

',

Thermal Processes
wt.-:

81

Figure 4.4

Product distribution obtained in the P E pyrolysis at diferent temperature^:^^ methane ( I ) , hydrogen ( 2 ) , ethylene ( 3 ) ,propylene ( 4 ) , cyclopentadiene ( 5 ) , benzene ( 6 ) , toluene (7), benzene + toluene ( 8 ) . (Reprinted from J. Anal. Appl. Pyrof., 8, W. Kaminsky, page 439. 0 1985, with permission from Elsevier Science)

ethylene. Thus, a selectivity of 3 1.1 wt% towards the monomer was obtained at 790 "C. The product distribution obtained in the pyrolysis of two types of polyethylene with different degrees of branching has been investigated by Conesa et a1.,26also using a sand fluidized bed reactor at temperatures between 500 and 900 "C and different residence times. The main products observed in the gaseous effluent from the pyrolysis reactor were methane, ethane, ethylene, propane, propylene, acetylene, butane, butene, pentane, benzene, toluene, xylene and styrene. At the lowest temperatures investigated (500 and 600 "C), significant amounts of tars and waxes were detected in addition to gaseous products. It was observed that the more branched polyethylene yielded more aromatic compounds. Thus, benzene yields of over 20 wt% were obtained in the pyrolysis of branched PE at 800C. The authors suggest that the formation of large amounts of aromatic compounds may arise from the intramolecular abstraction of a hydrogen atom, giving a more stable aromatic ring. Williams and Williams27have studied the pyrolysis of both HDPE and LDPE in a fixed bed reactor. In each experiment the temperature was varied between 25 and 700C. The products were swept down through the bottom of the reactor by a nitrogen flow and separated into several fractions by condensation at different temperatures. Two main fractions were recovered as products from the HDPE and LDPE pyrolysis: gases with a yield of 15-17 wt% and oils with yields in the range 80-84 wt%. The gases were rich in ethylene, propylene and butene, with lower proportions of saturated hydrocarbons (methane, ethane, propane and butane). The oils produced were analysed by FTIR and GPC,

82

Chapter 4

being basically formed by aliphatic hydrocarbons with a certain proportion of olefinic groups and boiling points in the range 100-500C. Compared to the previous studies on fluidized bed pyrolysis, the high proportion of oils obtained in a fixed bed reactor is probably due to significant differences in the nitrogenpolymer contact and heat transfer rate between the two types of pyrolysis system. The composition of the oils formed in the degradation of PE have been investigated by different authors.21928 Figure 4.5 illustrates the GC analysis corresponding to the oil fraction obtained by LDPE pyrolysis at 420 "C.Most of the products present in this fraction lie in the range C5-C22.The signal of each carbon atom number is resolved into two main peaks: the corresponding linear alkane and 1-alkene. In some cases, a third small peak is also observed, indicating the presence of dienes. According to Wampler, l 2 the diene/alkane ratio increases with the pyrolysis temperature due to the enhancement of the C-H bond breaking, which generates elemental hydrogen and favours char formation. Recently, PE pyrolysis has been tested at temperatures between 650 and 850 "C in a novel reactor type, called a rotating cone reactor'9329 (see section 4 of this chapter for a description of the system). At 725"C, using a bench-scale reactor, 80% of the polyolefin is converted into gases whereas the remaining 20% accounts for a liquid fraction composed of wax-like products and aromatics. The gases in this case are rich in methane and C1-C.Q olefins. Likewise, using a pilot plant reactor, the optimum yield of gaseous alkenes was obtained at 750 "C, because at lower temperatures the intermediate liquid products are not converted, whereas high amounts of methane and aromatics appear at higher temperatures.

a-olefin n-pars
c7 c8 c9
(

c13c14

c16

c17 c18

I c19
c20

c21

Pt
I

2.0
Figure 4.5

4.0

6.0

8.0 10.0 12.0 Retention time (min)

14.0

16.0

18.0

GC analysis of the oils obtained by LDPE cracking at 420 "C, 90 min.21

Thermal Processes

83

Ultrapyrolytic conditions (high temperatures and short reaction times) have been applied by Lovett et al.30 for the decomposition of LDPE using an internally circulating fluidized bed reactor of hot sand. The residence time of the polyolefin was 600 ms. Operating at temperatures in the range 780-860 "C, this reactor allowed the gas yield to be increased to values of around 90 wt%. These gases were mainly ethylene, methane and propylene and minor proportions of butene, butadiene and ethane. Ethylene yields up to 37 wt% were observed at 865 "C, showing that PE ultrapyrolysis makes it possible to go back to the starting monomer to a certain extent. The authors have attributed the high gas yield obtained in this system to the use of very low residence times, which avoids secondary reactions of the primary pyrolysis products leading to aromatics and the production of condensates. Cozzani et al.31have also studied the pyrolysis of PE in a fixed bed reactor, varying both the temperature (500-800 "C)and the residence time. In contrast with the previously described studies, these authors take into account three product fractions: gases and tars, which are swept out of the reactor by a nitrogen stream, and a char or coke, which remains in the reactor as a solid residue. The char yield was found to increase with the pyrolysis temperature. Thus, while at 500C char formation was negligible, at temperatures in the range 700-800 "C the char yield reached values of over 20 wt%. According to the authors, the char originates from a secondary process involving the cracking of the previously formed tars. This is the reason for the absence of solid residues after complete PE conversion when the pyrolysis is carried out in systems that do not allow tar-cracking reactions to take place, as is the case in vacuum or flash pyrolysis. A further has focused on determining the properties and possible applications of the coke obtained by PE pyrolysis. SEM microphotographs of the coke formed by PE pyrolysis at 800C showed that, in contrast with the filamentous coke obtained at 500 "C,the coke formed at the higher temperature has a spherical structure, which suggests that it is formed through a gas-phase mechanism. The reactivity of this coke was characterized by TG analysis in oxygen and carbon dioxide, concluding that PE coke exhibits a reactivity slightly lower than that of coal chars and similar to that of diesel soot. Another factor that strongly influences the pyrolysis of PE is the presence of steam.33i34 Figure 4.6 shows the changes observed in the product distribution of PE pyrolysis at 800C in a fluidized bed reactor as the steam concentration is increased.33 Significant improvements in the ethylene and propylene yields are produced by steam cracking, whereas the aromatic yield is greatly reduced and the formation of char is suppressed. Moreover, some carbon oxides are detected in the gases, which indicates that steam is not chemically inert but that some gasification reactions are also produced. These positive effects of the presence of steam in the thermal degradation of polyolefins has led to the development of commercial processes for the production of olefins by steam cracking.35936 In contrast with these previous pyrolysis studies that try to maximize the gas yield, thermal cracking of PE at low temperatures is usually aimed at the production of waxy oil fractions. It was found that thermal degradation of PE

84

Chapter 4

50

KplV.1

too

Figure4.6

Variation of the product distribution in the PE pyrolysis at 800C in the ethylene, A propylene, benzene, presence of steam:33 0 carbon residue, x hydrogen. (From Angew. Chem., Int. Ed. Engl., 1976, 15(11), page 660, Wiley-VCH Verlag GmbH)

begins at temperatures as low as 350C, leading to a reduction in the polymer molecular weight. More extensive degradation yields hard, semi-solid and liquid waxy products. An increase in the unsaturation of the products occurs as the PE thermal cracking progresses. In fact, PE thermal decomposition has been used in the past as a process for the synthesis of different commercial waxes from the virgin p ~ l y m e r . ~ ~ . ~ ~ Recently, several processes have been proposed for the production of paraffin have waxes by the thermal cracking of polyethylenic wastes.3941 Chaala et al.41 investigated the thermal decomposition of electric cable wastes at 450 "C under vacuum. These wastes contained about 9 1 OO / of crosslinked polyethylene. The products were removed from the reactor by a vacuum pump and separated by condensation at different temperatures into six fractions with the following yields: gases (4.1 wt%), soft wax (33.7 wt%), hard wax (37.6 wt%), light liquid phase (1.2 wt%), heavy liquid phase (2.0 wt%) and solid residue remaining in the reactor (21.4 wt%). The latter consisted mainly of inorganic fillers and the carbon black present in the starting wastes, hence it cannot be considered a product derived from PE. The hard wax was mustard in colour with a melting point of 72"C, while the soft wax was brown, with a melting point of 67C. These waxes were characterized by FTIR and N M R measurements, showing that they were made up of slightly branched chains with side alkyl groups and a certain content of double bonds. The authors proposed several possible applications for these waxes: mixing with commercial petroleum wax, compatibilizer for polyolefin blends, and high value applications in polishes, carbon paper, electrical insulations, inks, greases, lubricants, etc.

Thermal Processes

85

Polypropylene
Polypropylene is a polyolefin found in high concentrations in the plastic waste stream. Of the different types of PP, isotactic polypropylene is the one most widely used on a commercial scale and so is the type predominant in plastic wastes. Compared to PE, the backbone of the PP molecule is characterized by the presence of a side methyl group at every second carbon. This fact implies that half of the carbons in a PP chain are tertiary carbons and so, as a consequence of their higher reactivity, PP is thermally degraded at a faster rate than PE. Thus, as can be seen in Figure 4.7, the PP weight loss in TGA measurements starts at a lower temperature compared to both HDPE and LDPE. Random chain scission of polypropylene produces both primary and secondary radicals. Subsequently, tertiary radicals are formed by intramolecular radical transfer reactions. b-Cleavage of these tertiary radicals may lead to both olefinic and saturated branched fragments. In addition, depolymerization reactions to yield propylene may also take place from primary and secondary radicals. A detailed description of the PP thermal decomposition mechanism has been given by Kiran and Gillham.42 These authors conclude that in the degradation of isotactic polypropylene, intramolecular radical transfer to the 5th, 9th and 13th carbon atoms in the secondary macroradicals and to the 6th, 10th and 12th carbon atoms in the primary macroradicals account for the main products of the decomposition. The latter are proposed to consist of a series of highly branched alkanes and 1-0lefins. Likewise, Westerhout et aL2 have recently reviewed the kinetic models and parameters reported in the literature to describe PP thermal degradation, mainly those derived from TGA measurements.

I
100
80
n

60

c ,

.a

40

20 0

I .
100 200

I . I I 300 400 500 600 700


1 .

Temperature (C)
Figure 4.7

TG analysis o f PP in a nitrogen atmosphere.

86

Chapter 4

According to Figure 4.7, volatile products can be obtained by thermal cracking of PP at temperatures below 400C. Thus, Tsuchiya and S ~ m i ~ ~ carried out PP thermal degradation at temperatures in the range 360400C under vacuum. Gases with a high content of propylene were obtained but heavier hydrocarbons, mainly pentane, were also produced. Kiang et ~ 1have . studied ~ ~ the thermal cracking of both isotactic and atactic propylene, observing that isotactic PP degrades faster than atactic PP. Conversions of the isotactic PP of over 80% were obtained at 414 "C in less than 20 min, which leads to the formation of gases and liquids in an approximate ratio of 30/70 wt%. Whereas the gases contained a high proportion of propylene, the liquid fraction consisted mainly of branched olefins such as 2-methyl- 1-pentene, 2,4-dimethyl- 1-heptene and 2,4,6-trimethyl- 1-nonene. PP pyrolysis at temperatures above 600C has been carried out in reactors and systems similar to those previously described for the pyrolysis of PE. Using a fluidized bed reactor for PP decomposition at 740"C, Kaminsky et al.34 obtained a 57.3 wt% yield of gases containing mostly methane, ethylene and propylene. The liquid fraction, with a yield of around 40 wt%, was formed in similar proportions by aliphatic and aromatic compounds. Significant yields of benzene and toluene were obtained. Westerhout et ~ 1 . have ' ~ studied ~ ~ ~ PP pyrolysis in rotating cone reactors. At 750 "C gases with a yield around 95 wt% were produced, comprising mainly propylene, ethylene, butadiene and methane. On the contrary, Williams and Williams,27working with a fixed bed reactor and a pyrolysis temperature of 700 "C, obtained a gas yield of just 15 wt%. As in the case of PE, the high proportion of oil fraction produced with this system is probably due to a slower heating rate of the reaction mass and a poorer contact with the gas phase compared to pyrolysis in fluidized or circulating bed reactors.

Polystyrene
Polystyrenic plastics constitute a significant part of industrial and household wastes. As in the case of polypropylene, half of the carbons in the polystyrene chain are tertiary due to the presence of side benzylic groups. Therefore, thermal PS degradation also occurs at relatively low temperatures. TG analysis of PS in a nitrogen flow (Figure 4.8), shows that the thermal cracking of this polymer with formation of volatiles starts at temperatures around 350 "C. PS thermal degradation also proceeds through a free-radical mechanism initiated by random chain scission. Primary, secondary and tertiary radicals are involved in a series of transformations, mainly hydrogen transfer reactions and p-scissions, to yield the final degradation products. Detailed descriptions of the PS thermal degradation mechanism can be found in the literature.4549 A number of studies have been performed on the kinetics of PS thermal decomposition mainly based on TGA measurement^,^'-^^ which were recently summarized by Westerhout et a1.20Likewise, a model has recently been developed to describe the MWD evolution during PS thermal d e g r a d a t i ~ n . ~ ~ The formation of radicals during PS thermal degradation has been confirmed by Carniti el d7 through ESR measurements within sealed tubes. Figure 4.9

Tliernzal Processes

87

I w

.-

lg

1
Temperature ("C)

Figure 4.8

TG analysis of P S in a nitrogen atmosphere.

illustrates the changes in the ESR signal intensity (proportional to the free radical concentration) with time for different degradation temperatures in the range 350420C. The intensity of the ESR signals increases rapidly at short times, which accounts for the initiation step of the mechanism, the rate of radical formation being proportional to the temperature. At 420 "Cthe intensity

Figure 4.9 Changes in the intensity of the ESR signals during P S cracking at diferent temperature^:^^ 0 350"C, 360"C, 0 380"C, A 400"C, 0 420C. (From J. Polym. Sci. Polym. Chem., 1989,27, page 3865, Wiley-VCH Verlag GmbH)

88

Chapter 4

passes through a maximum, denoting that at that point the termination step becomes more effective. In most of the works reported on PS thermal degradation the major product obtained is the starting monomer. This fact is valid for both low and high temperature degradation. Therefore PS is one of the few polymers that can be thermally depolymerized. Schroder and Ebert58 have studied the thermal degradation of PS at temperatures of about 300 "C. From the evolution of the product distribution with time, the authors identified three types of products:
0

Primary products, which are formed almost instantaneously and whose concentration decreases with the degradation time. Examples include styrene, diphenylbutene, triphenylhexene and other trimers. Stable final products, whose concentrations increase with the reaction time u p t o a certain limit. Toluene, ethylbenzene, cumene and triphenylbenzene are the main species that follow this trend. Intermediate products, which have a maximum in their concentration, approaching zero at high conversions. The compounds exhibiting this behaviour are a-methylstyrene, diphenylpropane and diphenylbutane.

Zhang et al.59 have reported that the thermal degradation of PS in a fixed bed reactor at 350 "C under nitrogen flow leads to a styrene selectivity of 70 wt% with a conversion of > 80%. Other products identified were the styrene dimer, a-methylstyrene, toluene and ethylbenzene. A different product distribution was observed by Carniti et a1.60in the thermal degradation of PS at temperatures in the range 360-420 "C within glass tubes sealed under vacuum. In this case, the main products observed were toluene and ethylbenzene, instead of styrene. Figure 4.10 shows the changes in the styrene/ethylbenzene ratio with the reaction time at different temperatures. The decrease in this ratio with both the time and the temperature suggests that styrene is initially formed, which is then further converted into ethylbenzene. This secondary transformation takes place because the sealed tubes d o not allow styrene and volatile components to leave the reaction zone, in contrast with open reaction systems. Williams et a1.6' have analysed in detail the composition of the oils produced during PS degradation at 500 "C in a batch reactor with secondary cracking of the volatile products at 500, 600 and 700 "C. The main products obtained were styrene, benzene, xylene, toluene, styrene oligomers and a number of alkylated styrene derivatives. Moreover, significant concentrations of a variety of polyaromatic hydrocarbons (PAHs) were also detected: phenanthrenes, fluoranthenes, benzopyrenes, chrysene, etc. Most of these components are known to be carcinogenic and/or mutagenic agents, hence their presence may limit the possible application of the PS pyrolysis oil directly as a fuel. The concentration of PAHs increased with the temperature of the secondary cracking. The authors propose that PAHs are mainly formed by Diels-Alder reactions starting from styrene and styrene oligomers. Murakata et al.62,63 have studied the effect of different solvents (l-methyl-

Thermal Processes

89

100

200

xxl t (mid

400

500

Figure 4.10 Evolution o f the styrenelethylbenzene ratio with the time during the P S cracking at diferent temperatures:60 0 360C, A 380"C, 0 400"C, 0 420C. (Reprinted with permission from P. Carniti, P.L. Beltrame, M. Armada, A. Gervasini and G. Audisio, Ind. Eng. Chem. Res., 30, 1624. 0 1991 ACS)

naphthalene, decalin, tetralin, phenol and 2-naphthol) on the thermal degradation of PS in the temperature range 250-450 "C.The experiments were conducted in a stirred autoclave. According to the authors, thermal degradation of polymers in solution can be advantageous because it is not hindered by the mass and heat transfer limitations present when thermally treating pure polymers. The presence of the solvent affected both the PS conversion and the product distribution. A significant reduction in PS conversion was observed in the presence of tetralin, which is a solvent with a hydrogen donating ability. This result was explained by hydrogen abstraction from the solvent to intermediate polymer radicals, resulting in an interruption of the propagation reactions of PS depolymerization. The highest conversions were obtained with the phenolic solvents, thought to be because they promote the degradation of PS through hydrogen abstraction by phenoxy radicals. In addition, in contrast with conventional PS degradation, relatively low yields of styrene were obtained, the main products observed being ethylbenzene and toluene. These differences are probably due to the fact that the initially formed styrene remains in the reaction system, which favours its further conversion into the finally observed products. Degradation in the presence of solvents has also been applied to the conversion of other styrenic polymers, such as poly(p-methylstyrene) and poly(styrene-ally1 Figure 4.1 1 shows the temperature dependence of the conversion of poly(p-methylstyrene) when using phenol, 1-methylnaphthalene and tetralin as solvents. In this case, tetralin leads to the greatest degradation below 370"C, whereas the order is reversed above that temperature. These results indicate that the effect of the solvent in the degradation of styrenic polymers is strongly influenced by the temperature.

90
100 I

Chapter 4

300

350 Temperature

400

('C)

Figure 4.1 1 Temperature dependence of the poly (p-methylstyrene) conversion in the presence of diferent solvents:64 phenol, 0 I-methylnaphthalene, 0 tetralin . (Reprinted from Polymer, 34(7), T. Murakata, S. Wagatsuma, Y. Saito, T. Suzuki and S. Sato, page 1431. 0 1993, with permission from Elsevier Science)

A new approach to PS thermal degradation has recently been developed by Beltrame et af.66 based on PS conversion in the presence of water. The experiments were performed at temperatures between 300 and 350 "C in closed autoclaves under pressure and an argon atmosphere. From changes in the product distribution with time, the intermediate nature of styrene was confirmed, which was then further converted into toluene and ethylbenzene. Thus, styrene can be selectively obtained with this system only at short reaction times. The presence of water caused an increase in both the yield of light products and the styrene selectivity. The authors propose that water lowers the rate of both secondary cracking reactions and crosslinking reactions leading to solid residues. Due to its hydrogen donor character, water appears to prevent the intra- and intermolecular hydrogen transfers and the consequent shifts of the radicals along and between polymeric fragments. High yields of the starting monomer are also obtained in PS pyrolysis. Ericsson' reported that PS degradation in a commercial pyrolyser leads to styrene yields of over 60%. An increase in the pyrolysis temperature caused a lowering of the monomer production. The highest styrene yield (around 85 wt%) was obtained at 600 "C. Similar results were obtained by Audisio and when studying PS pyrolysis between 600 and 750 "C,styrene yields in the range 70-90% being observed. In contrast with the degradation of PS at low temperatures, significant amounts of benzene were detected in the pyrolysis product. The styrene yield was also correlated with the PS molecular weight. Longer PS chains led to higher proportions of styrene. The authors propose that styrene is formed by fl-scission reactions whereas toluene, ethylbenzene and a-methylstyrene are formed essentially by intramolecular hydrogen transfer.

Thermal Processes

91

The high selectivity to styrene during PS pyrolysis does not appear to depend reported styrene yields of 79.8 on the type of reactor used. Thus, Sinn et and 7 1.6% when the pyrolysis was carried out in a sand fluidized bed reactor at 640 and 740"C, respectively. Styrene yields of up to 92% have recently been reported by Lovett et uL3' during PS pyrolysis in a microreactor at high temperature (965 "C) and very short residence times (500 ms).

Polyvinyl Chloride
Polyvinyl chloride is a polymer with a wide range of commercial applications. However, its use has been the subject of great controversy in recent years due to its high chlorine content. Approximately 56 wt% of the polymer is HCl, which is released at relatively low temperatures, creating toxic and corrosive conditions. In addition, for many PVC applications it is necessary to incorporate significant amounts of other compounds (plasticizers, stabilizers, antioxidants, etc.), which will also be present in PVC wastes. Interactions between these agents and PVC during degradation have been described in the literature, and may lead to the formation of various unwanted Cl-containing organics. Therefore, treatment of PVC-containing wastes is not an easy task. PVC is a polymer characterized by its low thermal stability, hence numerous studies have examined its thermal b e h a v i ~ u r It . ~has ~ ~been ~ ~ observed that PVC is more susceptible to dehydrochlorination than C1-containing low molecular weight compounds. The presence of different types of defect sites in the polymeric chains has been proposed to explain the low temperature release of HCl from PVC:74-79tertiary and allylic chlorines, internal unsaturations, head-to-head configurations, etc. Likewise, the low thermal stability of PVC has been related to the presence of residues of certain additives used in the polymerization reaction. The major role of the stabilizers added to PVC is to decrease the rate of dehydrochlorination. As a consequence of the low temperature removal of HC1, thermal decomposition of PVC is a two-step process: dehydrochlorination of the polymer to form a polyene macromolecular structure followed by cracking and decomposition of the polyene. Figure 4.12 illustrates a typical T G analysis of PVC,80 showing clearly two steps of weight loss which correspond to each of the decomposition stages. The first transformation occurs in the temperature range 200-360 "C with a weight loss of 58%, which perfectly correlates with the theoretical HCl content. Small amounts of benzene and other hydrocarbons have also been detected in the gases generated in this step. Chemical analysis of the residue after treatment at 375C found no elemental chlorine. This result indicates that most of the C1 present in the starting polymer can be removed by a low temperature treatment, which is the basis of many of the processes developed for the treatment of PVC' proposed that the presence of two containing plastic wastes. Wu et ~ 1 . ~have inflexion points in the first weight loss step of PVC suggests that the first transformation is not a simple reaction but is really made up of two decomposition processes related to the existence of configurations in the polymer chains

92

Chapter 4

0-

-20-40Tonrat = 689.1 K Ttop ~ 7 1 9 . 2 K

-60-80 -

Tonsat 539.0 Ttop = 559.4 K


1

J,
.075

273

373

473

573 T (K)

673

773

Figure 4.12 DSC ( I ) and TGA ( 2 ) curves of PVC under a nitrogen atmosphere ( 5 K/ min).80 (From Can. J. Chem. Eng., 1994, 72, page 644, Canadian Society for Chemical Engineering)

with different thermal stabilities (head-to-head and head-to-tail configurations). As HC1 is being removed, the remaining residue undergoes a series of colour changes (white, reddish-brown, dark blue and black) which confirms the progressive formation of a chromophoric conjugated polyene structure. The second weight loss begins at a temperature of about 380 "C and it extends up to temperatures slightly above 500 "C. The volatile products formed at this stage are a complex mixture of aliphatic, olefinic and aromatic hydrocarbons. M,ore than 170 products have been identified in the gases evolving during the second step.*' The weight loss does not go to completion but a solid residue is obtained after heating at 6OO0C, accounting for about 10 wt% of the starting PVC. The proportion of this char depends on factors such as the final temperature and the heating rate during the TGA experiment. It has been suggested that the char is formed from the polyene structure through a reaction coupled in parallel with the polyene conversion into volatiles. Mechanistic information on the PVC thermal degradation can also be derived from the differential scanning calorimetry (DSC) curve shown in Figure 4.12. Two large endothermic peaks are observed at 286 and 446 "C. The first indicates that the dehydrochlorination step, involving the cleavage of the C-Cl bond, is an endothermic reaction. The appearance of a small exothermic peak at 358 "C has been attributed to the cyclization and crosslinking of the polyene structure. Several authors have proposed that the dehydrochlorination step of PVC involves autocatalysis by the released HCl.82-84This phenomenon has been demonstrated by Pate1 et al.,84who studied the degradation of PVC samples previously treated with HCl gas at different temperatures. Figure 4.13 compares the degree of dehydrochlorination at 150 "C for two PVC samples, one of them having been previously subjected to HC1 treatment. Dehydrochlorination of the

93

0.0

Figure 4.13 Degree of dehydrochlorination of PVC at 150 "C as a function of time: 0untreated PVC, HCl treated PVC.84 (From J. Appl. Polym. Sci., 1992,46, page 179, Wiley-VCH Verlag GmbH)

HC1-treated PVC proceeds to a larger extent and at a higher rate than in the case of the untreated PVC. Comparing results at different temperatures, the authors observed that the activation energy of this process was 20% lower for the treated PVC, confirming the autocatalytic role of HC1 during PVC degradation. Further experiments showed that the autocatalysis also occurs with the HC1 normally formed during the degradation. Thus, a decrease in the thickness of the PVC sample lowers the dehydrochlorination rate, as it favours the diffusion and escape of the HC1 formed. The formation of diene sequences after dehydrochlorination has been confirmed by UV-visible measurements. Figure 4.14 compares the UV-visible spectra of virgin and HC1-treated PVC after degradation at 120C for 9 h.s4 The higher absorbance obtained with the treated PVC sample reflects its higher degree of dehydrochlorination. The arrows indicate the wavelength of maximum absorption corresponding to specific conjugated diene sequences. Preferential absorption at these wavelengths is observed in both samples, especially in the HC1-treated PVC, showing the formation of these diene structures. Sinn et al.33 carried out PVC degradation in a sand fluidized bed reactor at 740 and 845 "C. In addition to about 56 wt% of HCl, other products obtained were methane, ethylene, benzene, toluene and naphthalenes. Around 8.8 wt% of carbon residue was also produced. Oudhuis et aL8' have studied the effects of some of the additives typically added to PVC on its thermal degradation. They used three different PVC

94
0.3

Chapter 4

0.2

0
[n

0.1

0.0 280

320

360

400

440

(nm>
Figure 4.14
UV- Vis spectra o f PVC degraded at 120 "Cfor 9 h: HCI treated PVC (A), untreated PVC (B). The arrows indicate the wavelength o f maximum absorption corresponding to a given polyene sequence, whereas 3, 4, 5 , 6, 7, 8 are the number o f double bonds contained in the latter.84

samples as starting materials: a virgin pure PVC, a rigid PVC incorporating metal stabilizers and a flexible PVC prepared by addition of metal stabilizers and a plasticizer to virgin PVC. Figures 4.15(A), (B) and (C) show the DTG curves of these samples in different atmospheres. In the case of rigid PVC a certain influence of the metal stabilizers is observed, the PVC degradation being shifted towards higher temperatures. This is as expected, because PVC stabilizers are added to PVC to delay dehydrochlorination. For flexible PVC a new peak appears at low temperature due to the evaporation of the plasticizers, although no influence of the latter on PVC degradation is observed. Because the plasticizer is released prior to the start of PVC degradation, it may be possible to separate and recycle the plasticizer during the thermal processing of PVC wastes. On the other hand, for both rigid and flexible PVC, there is an increase in the amount of solid residue finally obtained, which has been assigned to a carbonizing effect promoted by the stabilizers. Similarly, Oudhuis et al.85 have also investigated the formation of HCl at constant temperature for virgin, rigid and flexible PVC, concluding that the stabilizers present in these materials decrease the rate of HCl formation, especially at lower temperatures. The authors also studied the possible formation of polychlorinated dibenzodioxins (PCDD) and polychlorinated dibenzofurans (PCDF) during the treatment of virgin and rigid PVC at 352C in an inert atmosphere. The oils produced consisted mainly of benzene, toluene, naphthalene and anthracene with a chlorine content below 0.01 wt% in the form of chlorobenzene and chlorotoluene, which clearly reduces the commercial value of this oil. Moreover, as can be seen in Figure 4.16, the PCDD and PCDF emissions during the isothermal treatment at 352 "C were below 200 ppt for both types of PVC. In contrast, when this treatment was carried out in an

Thermal Processes

95

-0.0006

A
I

Q I 0

-0.00 1 5

-0.0025 373

573

773

973

0.0005 I

B
-0.00 15

-0.0025' 373

'

'

'
97' 3

573

773

TIK

(0
J

-0.0005

v) 1

s
-0.0015

C
'
I

-0.0025 373

573 TlK

773

973

Figure 4.15 DTG curves of PVC in dflerent a t m o s p h e r e ~ (A) : ~ ~ virgin P V C , (B) rigid P VC, (C)$exible P VC. (Reprinted from J. Anal. Appl. Pyrol., 20, A.B.J. Oudhuis, P. De Wit, P.J.J. Tromp and J.A. Moulijn, page 321. 0 1991, with permission from Elsevier Science)

oxidative atmosphere, the emission levels of these compounds increased by factors of between 10 and 1000. These results shows that no problems related to PCDD and PCDF emissions exist in the low temperature degradation of PVC. The possible interaction between PVC and several commercial plasticizers has also been investigated by Marcilla and Beltran.86987 Figures 4.17(A), (B) and ( C )show the TGA and DTG curves corresponding to plastisols formed by PVC and three commercial plasticizers (DBP, DOP and DIDP). The first weight loss

96
10000
c
\

Chapter 4

a a
C

0 .c
) 1

1000

I
4 5

100

10 6 7 8 4 5 6 7 number of chlorine atoms


Flexible P V C

=
Figure 4.16

Virgin PVC

PCDD and PCDF emissions during the treatment of virgin andflexible P V C at 352 OC.*' (Reprinted from J . Anal. Appl. Pyrol., 20, A.B.J. Oudhuis, P. De Wit, P.J.J. Tromp and J.A. Moulijn, page 321. 0 1991, with permission from Elsevier Science)

corresponds to the plasticizer removal, whereas the second is due to PVC dehydrochlorination. Compared to the evolution of the plasticizer alone, in all cases a delay is observed when it is included in the plastisol. At the same time, the presence of the plasticizer causes a narrowing of the DTG peak corresponding to the resin. The authors propose that the plasticizer interacts more strongly with those PVC fractions of lower molecular weight and crystallinity. In a recent work, Gupta and Viswanath88 have investigated the role of different metal oxides in PVC thermal degradation. The authors propose the following mechanism to explain the reduction in the release of HCl and the formation of metal chlorides. The first step involves the formation of chlorine free radicals, which subsequently replace oxygen from the metal oxide to form metal chloride and oxygen free radicals. Finally, the latter abstracts hydrogen from PVC to form water. In most cases, the presence of metal oxides affects the PVC degradation rate. V205,Zr02, Cr203, Fe203,Moo3 and Ce02delayed the PVC dehydrochlorination, whereas SnO2, Ti02, Sb203, Cu20, CuO and A1203 promoted this reaction. The possible contribution of phenolic antioxidants, frequently added to PVC, to the formation of chlorinated phenols during PVC pyrolysis has recently been investigated by B l a ~ s o using ~ ~ Irganox 245 as a model PVC additive. When PVC and Irganox 245 were pyrolysed together at temperatures of 600-800 "C, the resulting product distribution indicated the existence of significant interactions between the decomposition of both materials. Thus, Irganox hinders the formation of aromatic products from PVC degradation, confirming its role in the deactivation of radicals, whereas thermal fragmentation of Irganox is promoted by the HCl evolving from PVC. This effect is even more pronounced when copper or iron are added to the mixture. On the other hand, different

Thermal Processes
11-

97

-- 0.05
0.5 - -

0.05

0.5 _-

-I

0-

-0.05

14 100

I
150

200

250

300

350

400

Tmperature ("c)

Figure 4.17

TG and DTG curves of plastisols formed by PVC and three commercial plasticizers (dotted line: experimental, continuous line: theoretical):g6 (A) PVC + DBP, (B) PVC + DOP, ( C ) PVC + DIDP. (Reprinted from Polym. Degrad. Stab., 53, A. Marcilla and M. Beltran, page 261. 0 1996, with permission from Elsevier Science)

Relative amount
400

300

200

100

0 600C
Figure 4.18

700C

800C

Relative amounts of chlorophenolsformed by pyrolysis of Irganox with cupric ~hloride.'~ (Reprinted from J . Anal. Appl. Pyrol., 40-41, M . Blazso, page 69. 01997, with permission from Elsevier Science)

98

Chapter 4

chlorinated phenols were also detected when Irganox was pyrolysed in the presence of PVC, CuC12 or FeC13. Figure 4.18 illustrates the chlorophenols detected in the pyrolysis of Irganox with cupric chloride, which are probably formed by reaction of the cupric chloride with the phenolic radicals created by thermal cleavage of the antioxidant.

Other Plastics
This section briefly describes the thermal behaviour and conversion of other plastics, including materials such as PET, that are condensation polymers. As described in Chapter 2, condensation polymers are best depolymerized by chemolysis. However, the knowledge of both their thermal stability and the products derived from their thermal decomposition is of interest because in many cases they are present as contaminants in wastes containing addition polymers. Polymethyl methacrylate is one of the few plastics that can be thermally depolymerized leading to high yields of the starting monomer. Kaminsky and Franck have studied the thermal conversion of PMMA in different types of reactors. Figure 4.19 compares the product distribution obtained in a fluidized bed reactor and in a molten salt bath of a KCl/MgC12 eutectic. While the PMMA decomposition in the molten salt bath is slightly selective towards the monomer, yields of methyl methacrylate of over 90% are produced in the fluidized bed reactor over a wide range of temperature. In further experiments
100

Monomer
80

Fluidized bed

60

Hy dr oca r bons
Monomer
20

CO
co2

It

Fused salt bath

5 50

600

650

700

7 so

Temp. C

Figure 4.19

Products obtained in P M M A pyrolysis at dierent temperatures in fluidized and molten bath s y . ~ t e m s . ~ ~ (Reprinted from J . Anal. Appl. Pyrol., 19, W. Kaminsky and J. Franck, page 3 1 1. 0 1991, with permission from Elsevier Science)

Thermal Processes

99

the authors investigated PMMA depolymerization in a larger scale fluidized bed reactor with longer residence times. They reported yields of methyl methacrylate above 98% at temperatures of 450 and 490C. However, when the temperature was increased to 590C, a drastic decrease in the monomer yield was observed due to the formation of a variety of products such as methyl acrylate, methyl isobutyrate, dimeric methyl methacrylate, methanol, etc. The liquid product obtained by low temperature decomposition of PMMA could be polymerized to fresh PMMA without any further purification step, even when using wastes of this polymer coming from automobile rear lights. The high selectivity towards the monomer obtained by PMMA thermal degradation explains why this treatment is currently being applied on a commercial scale for the conversion of PMMA wastes. Madras et aL9' have investigated the kinetics of PMMA degradation in solution at temperatures between 200 and 300 "Cwith toluene as solvent. They concluded that PMMA depolymerization into methyl methacrylate takes place by end-chain scission, with no random cleavage. The thermal degradation of PMMA has also been studied by Ericsson," who compared it with the decomposition of polystyrene and a copolymer of alternating methyl methacrylate and styrene (PMMAS). Figure 4.20 shows the first-order kinetic plots corresponding to the decomposition of these three polymers at 500 "C. According to these results, it is observed that PMMA degrades faster than PS and the copolymer. For the latter, the formation rate of the two monomers is practically

6
5

0.2

0.4

0.6

0.8

1.0

Time I s Figure4.20 First-order plots of the monomers formed in the thermal de radation at 500 "C of P M M A , PS and alternating P M M A S copolymer:' 0 methyl methyl methacrylate methacrylate from P M M A , 0 styrene from PS, from P M M A S , 1 styrene from P M M A S .

100

Chapter 4

the same, suggesting that the degradation occurs by end-chain cracking (unzipping). It can be concluded that the bond between methyl methacrylate and styrene is stronger than that between two methyl methacrylate molecules but weaker than the bond between two styrene units. The thermal degradation of PET has been studied by Oudhuis et al.85 using TGA experiments. The DTG curves of PET in argon show a peak around 420 "C whereas 82% of the initial mass is volatilized up to 500 "C. The products released were a complex mixture composed mainly of acetaldehyde, benzoic acid, ethylbenzoate and vinylbenzoate. Likewise, Williams and Williams27 have investigated PET pyrolysis up to 700 "C in a fixed bed reactor, three fractions being collected: gases, oil and char. Gases and oil accounted for about 80% of the starting polymer mass. The gases were mainly carbon dioxide, due to the presence of oxygen in the PET macromolecules, although minor amounts of methane and ethylene were also detected. FTIR analysis of the oil showed the presence of aldehydes, ketones, carboxylic acids, alcohols and aromatic rings, probably substituted with at least one of these functional groups. Elemental analysis of the char indicated the presence also of a significant amount of oxygen (6.73%) and ash (5.86%). K a r n i n ~ k y ~has '.~~ reported the product distribution obtained in the fluidized bed pyrolysis of different condensation polymers (polyesters, polyurethanes, polyamides, etc.). Polyester degradation led to 51% of gases, with a high proportion of CO and CO2, and 40% of oil rich in benzene, toluene and naphthalene, the formation of water also being detected. On the other hand, polyurethane and polyamide decomposition led to the formation of about 40% gases and 55% oil. In both cases, the gases obtained contained certain amounts of HCN. The thermal degradation of polyamide-6 has been investigated by Shimasaki et al.93by means of TGA measurements in a nitrogen atmosphere. This polymer was almost completely volatilized in the temperature range 360-450 "C. The main product was c-caprolactam, with yields up to 50%, although significant amounts of cyclohexanone, propionic acid, N-vinyl-2-pyrrolidine and ammonia were also detected. When the TGA experiments were performed with polyamide-6 and melamine mixtures, an interaction between the components was observed. Melamine, which is a typical halogen-free fire retardant, increased the activation energy of the polyamide-6 thermal degradation and changed the distribution of secondary products.

4 Thermal Conversion of Plastic Mixtures


This section describes the different processes and treatments developed for the thermal decomposition of complex mixtures of several types of plastic, which is the case when processing real plastic wastes. Because the general aspects related to the degradation of each individual polymer have been commented on in previous sections, this section will emphasize technical factors, such as description of the reactors and processes, including the pretreatments which are

Thermal Processes

101

necessary for the processing of plastic wastes, as well as the possible interactions which may occur when several plastics are simultaneously degraded.

Interactions Between Components During Thermal Degradation


Only a few studies have appeared with the aim of determining the possible presence of interactions and synergistic effects during the degradation of polymer mixtures and real mixed plastic wastes. In some cases the results and conclusions of these studies are contradictory. Thus, while Wu et al.94did not observe any interaction between the components during the pyrolysis of a mixture of HDPE, LDPE, PP, PS, ABS and PVC, other authors have observed significantly different results when degrading mixed plastics compared to the conversion of the individual polymers. Several of these works are commented on below. Williams and Williams2' have investigated the pyrolysis of a plastic mixture in a fixed bed reactor by heating up to 700 "C.The mixture consisted of HDPE (31.25'/0), LDPE (31.25%), PP (7.29%), PS (13.5%), PVC (11.46%) and PET (5.21%), which simulates the plastic fraction of MSW. Figure 4.21 shows the different fractions obtained in the degradation of this mixture, and compares them to the products derived from the pyrolysis of the individual polymers. The plastic mixture was decomposed mainly into an oil fraction with a yield of around 75%. A theoretical product distribution for the plastic mixture was calculated from the results obtained in the conversion of each individual polymer. Comparing the real and the theoretical yields, some interesting differences were observed. Thus, the HCl collected from the mixture was only one-third of that expected according to its PVC content, which may be due to the occurrence of organochlorine compounds in the oil fraction. Likewise, a significant increase in the amount of char was observed when converting the

80 70
n

0HDPE
LDPE PS PP

s
w

60
50

a
c ,

5
0

40

30
20
10

0
GASES

OIL

CHAR

HCI

Product
Figure 4.21 Comparison o f the products obtuined in the pyrol sis at 700C o f HDPE, L D P E , PS, P P , PET, P VC and CI plastic mixture.

102

Chapter 4

1 0

15

20

25

Carban number

15

5
Figure 4.22

-25

10

15

20

Carbon number
Product distribution per atom carbon number in the liquid fraction obtained f PE, PE+ PVC, and PE+ PET mixtures at by thermal degradation o 430 0C.95 (Reprinted from Polym. Degrad. Stab., 53, Y . Sakata, M.A. Uddin, K. Koizumi and K. Murata, page 1 1 1. 0 1996, with permission from Elsevier Science)

Thermal Processes

103

mixture of polymers. The analysis of the oil from the plastic mixture by FTIR indicated the presence of aromatic and oxygenated groups in much larger amounts than expected by simple accumulation of the oil fraction derived from each polymer. All these facts suggest that the primary products formed by degradation of each polymer may react with products from the decomposition of the other plastics present in the mixture, which results in significant modifications of the product distribution. Sakata et al.95 have recently studied the thermal degradation of PE + PVC and PE + PET mixtures using a batch reactor under a nitrogen atmosphere. Figures 4.22(A) and (B) illustrate the product distribution per atom carbon number corresponding to the liquid fraction obtained by treatment of these mixtures at 430"C, as well as the results corresponding to the individual polymers. In both cases, synergistic effects were observed. For PE + PVC mixtures, the addition of PVC to PE causes a decrease in the proportion of heavier products, whereas in the PE + PET mixtures, the incorporation of PET caused a decrease in the yield of liquid products and an increase in the formation of gases and residues. To explain these results, the authors suggest that the interaction between PE and free radicals produced from the decomposition of PVC or PET may promote PE degradation, leading to the formation of lighter products. Likewise, Blazso et al.96 have investigated the thermal decomposition of PE in the presence of several chlorine-containing polymers: PVC, poly(vinylbenzy1 chloride) and poly(ch1orostyrene). Significant changes were observed in the yield of PE degradation products when a few per cent of these polymers was added to the PE. Figure 4.23 shows that the yield of aliphatic products is strongly reduced at 600 and 800C by the presence of PVC. A decrease in the yield of aromatic and polyaromatic hydrocarbons was also observed, especially in the presence of PVC. Some chloroaromatic compounds were detected by FTIR analysis of the tar fraction. The authors propose that HCl released from the C1-containing polymers promotes initial PE chain scission at the weakest points but hinders p-scission of the macroradicals. The huge variety of materials found in plastic wastes may lead to unexpected phenomena when they are thermally treated. Thus, in a recent work, Sakata et have reported that, during the low temperature dehydrochlorination of municipal waste plastics, a sudden production of large quantities of wax-like hydrocarbons has often been observed. This is an undesirable phenomena as it may cause a blockage of the gas lines which exit from the dehydrochlorination reactor. Based on experiments on the thermal treatment of PVC and aluminium mixtures, these authors conclude that the reason for the spontaneous degradation is the formation of hot spots caused by the heat released in the reaction between the HC1 evolving from PVC and aluminium foil contained in the plastic wastes. The local increase in temperature produced by this reaction would promote the partial thermal cracking of the plastics to form waxy products. Accordingly, the authors proposed that the detection and removal of aluminium or aluminium compounds is absolutely vital in plants processing plastic wastes by thermal treatments in order to avoid this undesirable spontaneous

104
Alkadienes Alkenes Alkanes

Chapter 4

PE
PE

400

600

800

Pyrolysis temperature ("C)

Figure 4.23

Yield of aliphatic products f r o m the thermal decomposition of PE and PE i P VC mixtures at diflerent temperature^.^^ (Reprinted from J. Anal. Appl. Pyrol., 35,M. Blazso, B. Zelei and E. Jakab, page 221. 0 1995, with permission from Elsevier Science)

degradation. Likewise, it is suggested that similar reactions may take place between HC1 and other metals, such as iron, present in the plastic wastes. The thermal decomposition of automotive shredder residue (ASR) has been studied by Rausa and P ~ l l e s eby l ~ flash ~ pyrolysis in the temperature range 650850 "C. The main polymers present in ASR are polypropylene, polyurethanes, polyester, ABS and polyvinyl chloride. TG-FTIR measurements in helium show that HC1 evolution starts at 250"C, whereas between 340 and 550"C, in addition to various hydrocarbons, HCl, NH3, H 2 0 , C 0 2 and some SO2 evolve from the sample. Finally, in the range 700-850 "C most of the oxygen-containing compounds are decomposed, because the gases formed are mainly CO and CO2. Further ASR degradation experiments were carried out in a flashpyrolysis system at different temperatures and with a fixed residence time of 2 s. A detailed chromatographic analysis of the products released allowed the detection of six main types of organic substances: n-alkanes, n-alkenes, polyenes, heteroatom-containing compounds (0, N and Cl), mononuclear aromatics and polynuclear aromatics. The distribution of these fractions is strongly affected by the temperature. Alkanes and alkenes consist mainly of straight chain molecules. The polyenes detected were mainly butadiene, pentadiene and cyclopentadiene. Above 650 "C, there is a significant increase in the production of aromatic species. At 850 "C, polynuclear aromatics are the main components of the pyrolysate. The presence of phenol, benzaldehyde, aromatic nitriles, pyridine and chlorobenzene has been detected between 650 and 800C. The changes in the relative concentrations of these components with temperature is shown in Figure 4.24. All these results indicate that ASR pyrolysis above 650C causes the production of substances with a major environmental impact.

Thermal Processes
1000
5

105

800

2 Y
rd

600

400

200

650 "C

800

"C

850 "C

Figure 4.24

Products bearing heteroatoms ( N , 0, Cl) formed in the A S R pyrolysis at diferent temperature^:^' propylene nitrile ( 1 ) , pyridine ( 2 ) , chlorobenzene ( 3 ) ,phenol ( 4 ) , benzaldehyde (S), benzonitrile ( 6 ) , benzeneacetonitrile ( 7 ) . (Reprinted from J. Anal. Appl. Pyrol., 40-41, R. Rausa and P. Pollesel, page 383. 0 1997, with permission from Elsevier Science)

Processes for the Thermal Degradation of Plastic Wastes


A variety of processes and reactors have been developed for the thermal processing of plastic wastes during the last 30 years, although most of them only at laboratory or pilot-plant scales. Important factors determining the product distribution are the reaction temperature, the heating rate, the average residence time and the reactor type. Both heat transfer rate and residence time are closely related to the reactor type and design, the following reactors being the ones most commonly used for the thermal decomposition of plastic materials: stirred tanks, shaft furnaces, rotary kilns, fixed beds, fluidized beds, circulating bed reactors and screw extruders. One of the earliest processes of plastic thermal degradation was developed by Matsumoto (see ref. 33). It consisted of a two-stage process: the feed material was first treated at 300C in a melting vessel, which caused the release of HCl from the PVC present in the starting mixture, and subsequently the molten plastics were cracked in a second tank at temperatures between 400 and 500 "C, depending on the raw material. When a plastic mixture formed by LDPE (35.7%), HDPE (14.2%), PP (lO.8%), PS (16.9%), PVC (18.0%) and other components (4.4%) was treated according to this process, the global product distribution obtained comprised 11.9% of gases and 73.2% of oil in addition to 7.7% of HCl. Fluidized beds are one of the preferred reactor types used in recent years for the thermal conversion of plastic wastes, due to the enhancement of heat and mass transfer processes typically associated with these systems. Many studies in this area have been performed by Kaminsky et a1.24933790392,997100 since the 1970s at the University of Hamburg involving the construction and testing of several plants at different scales (laboratory, pilot plant and semi-industrial plant). Figure 4.25 shows a schematic diagram of the pilot plant divided into four

106

Chapter 4

EACTOR

SEPARATION

PRODUCT PREPARATION

L
Figure 4.25
Schematic diagram of the process developed at the University of Hamburg for the pyrolysis of plastic wastes in fluidized bed reactors.99 (Reprinted from J . Anal. Appl. Pyrol., 40-41, J. Kim, W. Kaminsky and B. Schlesselmann, page 365. 0 1997, with permission from Elsevier Science)

zones:99 feeding and regulation of the fluidizing gas, fluidized bed reactor, separation of the produced fractions and product purification. Figure 4.26 shows in detail the reactor, which includes a fluidized sand bed heated by several radiating fire tubes.24 These tubes are internally heated by burning propane or pyrolysis gas. An inert gas or, preferably, the same pyrolysis gas is used as fluidizing agent, previously heated to 400C. Plastics can be introduced into the reactor through a double flap gate or by means of a screw. One of the main advantages of this system is the possibility of feeding large pieces of the starting waste materials, which eliminates the need to carry out costly operations for reducing the size of the raw plastics. Lime is usually added to the plastic wastes to capture the HCl formed from PVC. The product gases emerging from the fluidized bed are first separated from residual carbon and fine dust in a cyclone and are then cooled to obtain the final oil, gas and tar fractions. The fluidized bed plants developed by Kaminsky et al. have been used for the conversion of a variety of plastic mixtures. In a recent work,99 the results obtained in the conversion of two different plastic wastes at temperatures of 638, 690 and 735C were reported. The raw mixture consisted mainly of polyolefins (65-79%), polystyrene ( 6 3 0 % ) and polyvinyl chloride ( 4 5 % ) . The following fractions were derived from the pyrolysis of these materials: gases (3542.9%), oils (41-51.8%), residue (5.8-14.3%) and soot (2.2-5.4%0).The gases were mainly methane, ethylene and propylene, with a certain concentration of C O and C 0 2 due to the presence of oxygen in the feed material. At the lowest pyrolysis temperature the oil produced contains a high proportion of

Thermal Processes

107

Figure 4.26 Fluidized sand bed reactor o f the process developed at the University o f Hamburg:24 steel wall with fireproof bricks ( l ) , fluidized bed (2), tiltable grate (3), radiation fire tubes ( 4 ) , nozzles to remove sand ( 5 ) , flanges for observation (6, 8 ) , gas-tight lock (7), flange for repairs ( 9 ) , shaft f o r steel cord ( 10 ) . (Reprinted from J . Anal. Appl. Pyrol., 8, W. Kaminsky, page 439. 0 1985, with permission from Elsevier Science)

aliphatic hydrocarbons, while at higher temperatures the proportion of aromatics, mainly BTX, drastically increases. The styrene concentration in the oil is directly related to the PS content in the raw plastic mixture. The same trend is observed with respect to the nature of the residue, which changes from aliphatic to aromatic as the reaction temperature is increased. The soot was made up of sand and inorganic compounds, as well as aromatic solids and carbon black. After distillation, a portion of water was also obtained with a pH of 10, due to the presence of dissolved ammonia, probably derived from the occurrence of small amounts of nitrogen-containing polymers in the raw plastic mixture. An interesting conclusion of this work was that most of the chlorine was detected in the soot and the bed material. Nevertheless, the C1 content of the resulting oils was 20 ppm, still somewhat higher than the upper limit of 10 ppm imposed by petrochemical plants for the processing of raw materials. However, no chlorinated dibenzodioxins have been found among the chloro-organic compounds

108

Chapter 4

present in the pyrolysis oil. In a further work,'" the authors have shown that the C1 content of the oils could be reduced well below 10 ppm by using as raw material a plastic mixture previously depleted in PVC and through the distillation of the produced oils. The C1 was mainly concentrated in both pyrolysis and distillation residues. Paisley and Littlo2 have also developed a process for the thermal degradation of plastic mixtures in sand fluidized bed reactors. The polymeric materials are decomposed by heating to 800-900C at a rate of at least 500C s-l and residence times of less than 2 s, which allows relatively high monomer yields to be produced as secondary reactions, leading to tars and char, are prevented under such conditions. These extremely high heating rates are achieved by contacting the plastic with a fluidized bed of incandescent sand that passes through the reactor. The sand is first heated in a combustor at a temperature about 100-200 "C higher than the desired operating temperature in the reactor. The addition of calcium oxide is proposed to eliminate acid gases, mainly HCl, formed from the waste decomposition. The fluidizing agent is an inert gas or, preferably, steam, to enhance the monomer yield. When a plastic mixture composed of LDPE (49.2%), HDPE (34.5%), PS (11.3%) and PVC ( 5 % ) is pyrolysed in this system at 840 "C,the gases produced are made up of 42 mol% of ethylene, 26.9 mol% methane and 17.3 mol% hydrogen, with minor amounts of other components. Williams and Williarnslo3 have investigated the effect of temperature during the thermal degradation of a plastic mixture in a sand fluidized bed reactor, suggesting possible uses for the different fractions obtained. The starting plastic mixture comprised LDPE (31 Z % ) , HDPE (3 1.25%), PP (7.29%), PS (13.50%), PVC (1 1.46%) and PET (5.21 YO). An increase in temperature led to an increase in the production of gases with a reduction in the amounts of wax and oil fractions. The amount of char formed increased with temperature and was found amalgamated in the sand bed. The calorific value of the gases produced was around 40 MJ kg-', hence they could be used as fuel, although their high olefin content also makes them suitable for use in the synthesis of chemicals. The oil formed has a high concentration of aromatics, especially at high pyrolysis temperatures, so it could be used as a feed in a cracking unit to form ethylene. The authors propose that the wax fractions could be fed into catalytic cracking units to yield gasolines. Lovett et af.30have proposed the use of an internally circulating fluidized bed reactor to carry out the pyrolysis of plastic wastes. The reactor is designed to operate at temperatures in the range 600-950 "Cwith very short residence times, i.e. hundreds of milliseconds, and to achieve extremely high heating rates (> 50 000 "C s- I). Figure 4.27 shows a schematic diagram of the pilot plant including this reactor. The plastic particles are fed with the inlet riser gas (nitrogen or steam) and contacted with the hot sand which enters the base of the riser. At the top of the riser, the gases formed are quickly separated from the solids, the latter being recirculated to the annular bed. As described in the previous section, this reactor has been successfully tested for the pyrolysis of LDPE and PS polymers.

Carbon Absorption Bled

ICW Reactor and Furnace

Secondary Separator

7 w Gas Sample Point

Auxiliay Steam

___
Wet Test Meter

Exhaust

Feed
~

n '
Heater
Filter

- .- . - .- .boil or Plaslics!Oil ..........___ .Plastics Feed - - - - - _ _ - - Sleam Lines

Plastics Fecdcr

'

meter

Nitrogen Bottle

Pu'np

H c a y OiI/Plastics Reservoir

Figure 4.27 Schematic diagram of a pilot plant for the thermal conversion of plastic wastes with an internally circulatingfluidized bed rea~tor.~' (Reprinted with permission from S. Lovett, F. Berruti and L.A. Behie, Ind. Eng. Chem. Res., 36,4436. 0 1997 ACS)

110

Chapter 4

Another type of pyrolysis reactor based on a sand fluidized bed has recently been developed by Westerhout et af.'9.29 for the pyrolysis of plastic materials. Both bench-scale and continuous rotating cone reactors have been used for the thermal conversion of polyolefins. Schematic drawings describing the main parts of the continuous rotating cone reactor are shown in Figure 4.28. The system consists of three parts: the actual cone reactor, a bubbling fluidized bed reactor used to heat the sand, and a riser to transport the sand from the fluidized bed to the cone reactor. It is proposed that on an industrial scale air should be introduced into the fluidized bed to burn off the coke generated during the plastic pyrolysis and deposited onto the sand. The cone reactor is rotated at 600 rpm. The polymer is fed into the reactor with preheated sand, which is used to provide heat to the reactor, to increase the heat transfer rate, and to prevent the polymer particles sticking to the reactor wall. The sand and polymer particles are thoroughly mixed on the bottom plate of the cone, being transported upwards due to the centrifugal force generated by the cone rotation. Low polymer/sand ratios are used (1/62 to 1/108) to avoid large temperature gradients in the reactor. The polymer is pyrolysed as it is transported with the sand. The gas produced exits the reactor through two points in the cone (at the bottom and at half height), whereas the sand is transported from the cone
CFB fluidisation gas
Sand inlet
4

650 m m

BFB fluidisation
2lOm

CFB fluidisation gu inlet

Figure 4.28

Schematic diagrams of a pilot plant for the thermal conversion of plastic wastes with a continuous rotating cone reactor: (a) pilot plant, (b) cone reactor, (c) cone dimensions. l 9 (Reprinted with permission from R.W.J. Westerhout, J. Waanders, J.A.M. Kuipers and W.P.M. van Swaaij, Ind. Eng. Chem. Res., 37, 2316. 0 1998 ACS)

Thermal Processes

111

towards the bubbling fluidized bed by gravity. Nitrogen is used to transport the sand back into the cone reactor through the riser. According to the authors, this type of reactor has significant advantages compared to conventional pyrolysis systems: short gas and solid residence times, good polymer-sand mixing and no cyclone is required for the separation of the sand and the produced gases. This system was used for the pyrolysis of PE + PP mixtures with different proportions of both polymers. On comparison with the results obtained in the conversion of PE and PP individually, the authors did not observe any special effect derived from their coprocessing, which indicates that for reactors with short residence times, secondary reactions between products resulting from the degradation of each polymer are not significant. A sand fluidized bed reactor is also used for the thermal conversion of plastic wastes in the process developed by BP in collaboration with Hamburg This process is planned University and a group of affiliated to be commercialized soon, with a demonstration plant of 25 000 tonnes/year operational in 2001. The starting raw plastic mixture must preferably be made up of polyolefins, with limited amounts of polymers containing elements other than carbon and hydrogen. The plastic mixture is first shredded into 2 4 cm pieces, which can then be screw-fed into the reactor. It is remarkable that, in contrast with the earlier described studies using fluidized beds, in this case the polymer degradation proceeds at a relatively low temperature, mainly between 500 and 550C. Calcium oxide is added to the fluidized bed to trap the HCl released by the PVC present in the feed and nitrogen is preferentially used as fluidizing gas. Under these conditions only around 10% of gases are obtained, intended to be used mainly as a fuel to heat the fluidized bed reactor. Likewise, at these low temperatures the formation of aromatics and char is minimized, hence the products consist mainly of aliphatic hydrocarbons, which are separated into a number of fractions (heavy waxes, waxes, light waxes and oil) according to their condensation points. The major application envisaged for these waxy products is their use as feeds in steam cracking refinery units to produce ethylene and propylene, which can be used as feedstocks for the synthesis of fresh polymers. In such cases, the amount of PS present in the starting plastic waste must also be limited, to avoid the formation of significant amounts of aromatic hydrocarbons, which are less desirable in steam cracking units. It is proposed that the polymer cracking plant must be located near a refinery in order to improve the process economy. The Veba Oel pyrolysis process is based on the use of a rotary kiln reactor. lo' Figure 4.29 illustrates the basic units of this process which has been used for the treatment of various types of residues and raw materials. Figure 4.30 shows the flow diagram for its use in the processing of plastic wastes. The raw materials are fed to the rotary kiln via a screw feeder. The kiln is indirectly heated by gas burners, and the residence time can be varied by changing the inclination and rotation speed. The pyrolysis is carried out at temperatures between 650 and 850 "C. The volatile products formed are separated and recovered by a two-step condensation. The solid residue of the pyrolysis is discharged into two alternating containers. When treating plastic wastes, lime can be added to the pyrolysis

112
W4STE LlOUlD

Chapter 4
WITERlAL
SOLID

FLUE GAS

cw

Figure 4.29

(Reprinted from J. Anal. Appl. Pyrol., 25, H.P. Wenning, page 301. 0 1993, with permission from Elsevier Science)

Veba Oel process for the pyrolysis of wastes. * O7

1
Figure 4.30

Add
2.4

Hydrcchlork

rm

Schematic dia ram of the Veba Oel process applied to the conversion of plastic wastes. 807 (Reprinted from J . .Anal. Appl. Pyrol., 25, H.P. Wenning, page 301. 0 1993, with permission from Elsevier Science)

drum to capture the chlorine as calcium chloride. However, the presence of the latter in the coke produced would hinder its subsequent application as a fuel. Therefore, it is preferable to remove most of the chlorine in a previous extrusion step at 350 "C,which allows the C1 content to be reduced to < 0.1 wt%. In this case, the coke produced can be used in existing coal power stations. The oil obtained has a high content of unsaturated hydrocarbons, and so it is proposed that it is further upgraded into a synthetic crude oil by catalytic hydrogenation.

Thermal Processes

113

On the other hand, the non-condensable gases consist mainly of hydrogen, methane, ethane and ethylene, with minor amounts of C3and C4 hydrocarbons, in addition to pollutants such as NH3, HCN and HCl. The latter must be removed by various treatments. A different approach is used in the Parak process,1o8aimed at the thermal decomposition of scrap plastics into paraffin waxes in the range C18-C50. The raw solid waste is first cleaned of contaminants and shredded. In order to reduce the viscosity of the molten plastics, they are mixed with slack waxes or with a part of the previously obtained cracked product. Subsequently, they are thermally degraded in a two-step process: low temperature treatment in a stirred tank reactor (320-350 "C) followed by a cracking at higher temperature. It is claimed that the waxes produced are of a quality comparable with that of standard commercial products, a variety of possible applications being suggested: candlemaking, corrosion protection, cosmetics, pharmaceuticals, etc. This process is being commercialized through the construction of a 20 000 tonnes year- plant. Other interesting systems developed for the thermal decomposition of plastic wastes include solvent-promoted degradation and degradative extrusion processes. Coenen and Hagen"' have patented a process for the production of liquid hydrocarbons from polyolefinic wastes by treatment with an organic solvent at 3 10 "C under pressure. The preferred solvents are aromatic compounds (benzene, toluene, xylenes and ethylbenzene). Under these conditions, almost complete degradation of PE and PP was obtained after 4 h of reaction. The products were a mixture of solubilized hydrocarbons, which can be easily separated from the remaining solid residue. This residue accumulates most of the contaminants and inorganic compounds present in the starting wastes. Tsukagoshi et a1.I" have developed a number of continuous systems to promote plastic degradation by simultaneous compression and shearing actions over a wide range of temperatures. Shearing and compression effects are achieved by treating the plastics within externally heated screw extruders, although the compression action can be complemented by pistons. The combined effect of these actions, as well as that of the temperature, causes rapid decomposition of the plastic materials. Shearing and compression lead to a significant generation of heat due to internal friction, and so the energy to be externally supplied is reduced. Likewise, Michaeli and Lackner' have proposed that a degradative extrusion can be an interesting pretreatment process during the conversion of plastic wastes. Extrusion of the plastic mixture at temperatures in the range 300400C leads to a molecular breakdown of the plastics as well as to the release of HC1 by PVC decomposition. The possible further conversion of the resulting mixture in a second zone of the same extruder or in a second extruder is also considered by increasing the temperature or incorporating a reactive atmosphere. Figure 4.3 1 shows schematically a laboratory unit for the degradative extrusion of plastics. In addition to the temperature, the rotation speed of the screw is an important variable in this system as it determines the residence time and the shear rate. Degradative extrusion of different plastic mixtures causes a significant

'

'

114

Chapter 4
plastic mixtures

gas washing for neutralization

container

vacuum pump

Figure 4.31 Laboratory unit for the degradative extrusion of plastic wastes."

0 material5 BE, PP. PS. PET, PVCl 0 material 1 (PE. PP) A materid 2 (PE. PP, PVC)

shear rate [ S-'

Figure 4.32

Viscosity o f the products obtained b degradative extrusion o f plastic mixtures as a function of the shear rate. 6 ,

reduction in the viscosity of the molten mixture. Figure 4.32 illustrates the viscosity (measured at 200 "C) of the products obtained by degradative extrusion at 400 "C as a function of the shear rate, showing that the viscosity can be reduced by a factor of ten with this treatment. l 1 With respect to HCl removal, the authors observed that an effective dehydrochlorination was only achieved with extrusion temperatures above 350 "C, whereas treatment at 400 "C resulted in residual C1 contents of 200 ppm. Interestingly, the C1 content in the product after extrusion was not a function of the PVC concentration in the raw mixture, although it largely depends on the presence of other polymers, such as PS and PET. These results suggest that a part of the evolved HCl may react with products of the PS and PET degradation to form C1-containing organic compounds. A significant increase in the amount of residual Cl was also observed when extruding real plastic mixtures, which is related to the possible presence in this case of other Cl sources, mainly C1 salts, in addition to PVC. The possible thermal processing of plastic wastes directly in refinery units is also an alternative that has been considered by some companies.''29113 In this case, the plastic wastes must be dissolved and mixed, generally in low proportions, with petroleum fractions. For this approach to be feasible the most important criterion is that the addition of the plastic wastes must not lead to any significant variation in the process parameters of the refinery unit. The

Thermal Processes

115

processing of plastics admixed with petroleum fractions has been investigated in steam crackers, visbreakers, coking units, etc. One of the major limitations of this approach is the possible presence of undesirable elements, such as C1, in the starting plastic wastes that would then be introduced into the refinery streams. It has been suggested that cokers are more suitable for the refinery recycling of plastic wastes because they can tolerate a higher proportion of impurities.

5 Thermal Coprocessing of Plastic Wastes with Coal and Lignocellulosic Materials


In recent years the possible coprocessing of plastics and coal has started to be considered as an interesting alternative for the production of chemicals and fuels from these two types of raw materials. Most of the coal liquefaction processes are based on hydrogenating treatments, but the high cost of hydrogen is one of the factors that limits the economic feasibility of such processes. On the other hand, plastic wastes are usually hydrogen-rich materials, and so it has been proposed that plastics could act as hydrogen donors during coal thermal conversion. As will be discussed in Chapters 5 and 6, the possible coprocessing of plastics and coal is not limited only to thermal treatments, and catalytic and/ or hydrogenation alternatives are also being investigated. Palmer et aL.ll4 have studied the conversion of coal + PE and coal + PP mixtures in autoclave reactors under a nitrogen atmosphere. From the conversion of the individual components, the authors calculated the theoretical conversion of the mixtures assuming that no interactions between the two components take place. Figures 4.33(A) and (B) compare the predicted and measured conversions corresponding to a 1/1 coal + PE mixture at different

100 1

80

w
I -

..... ....
====~

....-....... El A
z
W

1
I

lW-q 90 ..........................................,.rr:: ................


80.................................
4 -

..........................................
............

7060-

0 UJ
CL:

70. 60. 50.......


...... 40.
...... 30.

0
0

> z

50-

.........................

...... 20.

..........

3 5 0 400 425 450 475 500 5 2 5 5 5 0

10

PYROLYSIS TEMPERATURE (C)

15 30 60 TIME (MINUTES)

120

Figure 4.33 Comparison of predicted and experimental conversions in the thermal treatment of coal -tP E mixtures: l4 (A) temperature efect, (B) time efect. (Reprinted from Coal Science, eds. J.A. Pajares and J.M.D. Tascon. 0 1995, pp. 1 5 2 3 526, ~ with permission from Elsevier Science)

116

Chapter 4

temperatures and reaction times, respectively. It can be seen that in the temperature range 425-550 "C there is a reasonable agreement between both conversions, indicating a negligible interaction between coal and PE. However, at temperatures below 425 "C the actual conversions are significantly higher than the predicted ones, showing the existence of a synergistic effect in the coprocessing of both materials. This synergism is also observed at short times in the experiments carried out at 425C (Figure 4.33(B)). Similar results and conclusions were derived from the coprocessing of coal + PP mixtures. Miura et a1.l l 5 have reported that one of the main problems in the coprocessing of coal and plastics is the poor contact between these two materials when they are simply mixed. The authors propose that this contact can be improved by premodification of both coal and plastics and with the addition of methanol as swelling agent. Therefore, polyethylene waste was thermally degraded at 400C to form a wax, whereas the coal was treated with H202 at room temperature to introduce COOH groups, which favours the swelling effect of methanol. In this case also, synergistic effects were observed by thermal treatment of coal +wax and coal + wax + methanol mixtures at temperatures between 368 and 764C. Thus, the conversion of a mixture with these three components led to a decrease in the char yield and an increase in the production of tars and gases, compared with the product distribution expected from the treatment of coal, wax and methanol individually. Likewise, Collin and Polaczek' l 6 have carried out the co-coking of coal with reactive pitches derived from plastic wastes. Direct co-coking of coal and plastics is not feasible because the polymers usually decompose at temperatures below the softening point of coal. The reactive pitches were prepared by thermal treatment at 370400 "C under pressure of mixtures composed of 50% plastic material and 50% normal coal-tar pitch. The addition of the reactive pitches obtained to weak-coking coals and coal blends resulted in improved coking properties, yielding cokes with increased mechanical strength and optical anisotropy. Sivakumar et al. l7 have investigated the coprocessing of lignocellulosic wastes, plastic wastes and coal by treatment with CO, water and alkali at 400445C under pressure. Due to the presence of large amounts of water, high pressures are reached at the operating temperature. This system has previously been reported to be successful in the conversion of coal and other organic materials into liquids. Treatment of individual plastics (HDPE, LDPE, PP, PS, PET and PVC) with CO/H20/alkali at 400 "C led to conversions usually above 90%. PS, PP, LDPE and HDPE were transformed in significant amounts into oils and gases, whereas the latter were predominant when treating PET and PVC. The products from the decomposition of HDPE, LDPE and PP were composed mainly of saturated compounds with little olefinic character. Interestingly, no Cl was detected in the oils in the PVC degradation. Coprocessing of PP and newsprint (biomass-derived product) yielded an oil with low oxygen content and high calorific value. Finally, coprocessing of coal and polyolefins with this system caused a reduction in the coal conversion and an increase of the asphalt fraction, indicating that interactions occur between the substrates.

'

Thermal Processes

117

In a recent paper, Di Blasi"' has simulated the pyrolysis of cellulosic and plastic wastes. The author concludes that the different reactivities and physical properties of cellulose and PE leads to large differences in conversions and degradation rates below 727 "C,and that the copyrolysis of these two materials is not beneficial. However, above 837 "Cthese discrepancies are reduced, and so some interactions during the simultaneous pyrolysis of both materials can be expected.

6 Thermal Conversion of Rubber Wastes and Used Tyres


In recent decades much interest has been generated in the reclamation of rubber wastes, mainly used tyres, by thermal decomposition. As tyres contain significant amounts of other components, in addition to rubber, the thermal conversion of tyres is a more complex process than the thermolysis of rubber alone. Tyres are composed of rubber and various reinforcing materials, such as textile cords, steel, fabric belts, steel-wire beads, etc. The type of rubber most widely used in tyre manufacture is styrene-butadiene copolymer (SBR) containing about 25 wt% of styrene although, depending on the final use of the tyre, other rubber polymers can also be added, such as natural rubber, cis-polybutadiene, etc. The rubber is previously vulcanized by treatment with sulfur, which causes crosslinking of the polymeric chains to increase the hardness of the rubber and prevent excessive deformation at high temperatures. Organosulfur compounds, zinc oxide and stearic acid are usually incorporated to accelerate and control the vulcanization process. The tyre also includes a large amount of carbon black to increase the strength and abrasion resistance of the rubber, as well as extender oils to improve its workability. A typical tyre composition, both by elements and by components, is summarized in Table 4.1.' l 9 During the last 25 years a variety of processes have been developed for the thermal degradation of tyres in order to recover valuable components from rubber wastes. 120-125 Stirred tanks, rotary kilns, fixed beds, fluidized beds and tray systems are examples of reactor types used for the thermal degradation of tyres. Several of these processes are now being used on a pilot plant and industrial scale. Basically, three fractions are derived from the thermal decomposition of tyres: gases, liquid oils and solid residues. In the past, the influence of the reaction conditions and the reactor type were studied in order to maximize

Table 4.1 Typical tyre composition


Component w t OO / 60-6 5 29-3 1 2-3 1-2 -2 -2

SBR or natural rubber Carbon black Zinc oxide Sulfur Extender oil Additives

118

Chapter 4

the production of one or several of these fractions. Typical distributions obtained in the temperature range 500-900 "C are as follows:' l 9 10-30% gases, 38-55% oils and 33-38% char. However, in recent years, significant research effort has gone into the improved characterization of these fractions and to find new and economically more interesting applications than their use as low quality fuels. The influence of the temperature in the range 300-700 "Cduring the pyrolysis using a fixed bed of used tyres has been investigated by Williams et reactor in a nitrogen atmosphere. An increase in temperature up to 600C caused a decrease in the yield of char, whereas it favours the production of gases and oil. Above 6OO0C, only slight changes were observed in the product distribution. The gases were composed mainly of hydrogen, carbon dioxide, carbon monoxide, methane, ethane and butadiene. As shown in Figure 4.34, the oil fraction has an apparent molecular mass of up to 1660 with a peak maximum at about 400. A wide variety of compounds were identified in the oil with a high proportion of aromatic hydrocarbons, including polyaromatic species. FTIR analysis showed the presence also of oxygenated species such as aldehydes, ketones and carboxylic acids. With respect to the solid residue, the authors observed an increase in the char surface area with both temperature and heating rate, values up to 60 m2/g being obtained. These surface areas are large compared with those of the raw carbon blacks used in tyre manufacture. However, the pyrolysis char contains a higher proportion of ash and has a larger particle size, which make this product an

(w t %)

IG
3
6
L

400

800 1200 1600 Apparent Molecular Mass ( M A 1 U n i t s )

Figure 4.34

Typical apparent molecular mass of a lyre derived (Reprinted from Fuel, 69, P.T. Williams, S . Besler and D.T. Taylor, page 1474. 0 1990, with permission from Elsevier Science)

Thermal Processes

119

alternative to carbon black only in undemanding applications. Chars from tyre pyrolysis contain about 15 wt% of minerals, mainly zinc oxide. Similar results on the product distribution and composition of the produced fractions have been reported by several authors using a number of experimental 19,127-130 A detailed characterization of the composition and properties of the different fractions obtained by tyre pyrolysis has been performed by Williams et ~ 1 . The l ~ major ~ application proposed for the gases generated is as fuel to supply a part of the energy input required by the same tyre processing plant. The oils are characterized by a high chemical complexity and calorific values in the range 35-40 MJ kg-', hence in many cases the major application proposed is their combustion to produce energy. However, the relatively low sulfur content of the oils (usually 1-2 wt%) and the presence of valuable chemicals make it feasible to process and upgrade the oils for use as a refinery feedstock. In this way, the major changes that most refineries are currently undergoing in order to increase their capacity for the processing of petroleum with increasing sulfur content is a factor that may also favour the further refinery upgrading of the oils derived from the thermal degradation of tyres. have investigated the thermal decomposition of tyres at Chaala and 500C under vacuum in the set-up shown in Figure 4.35, in which the major part consists of a horizontal reactor 3 m long and 0.6 m in diameter. The feed

NonwndcnsaOm

0 1 1

0 1 1

Figure 4.35

Pilot plant f o r the vacuum pyrolysis of used tyres.'29 (Reprinted from Fuel Process Technol., 46, A. Chaala and C. Roy, page 227. 0 1996, with permission from Elsevier Science)

120

Chapter 4

comprises cylindrical particles punched from the side wall of cross-ply tyres. The gases formed during the tyre thermal degradation are removed from the reactor by a mechanical vacuum pump, and are partially condensed in a system of two packed towers operating in series. The fraction of non-condensable gases is burned to provide part of the heat needed in the process. The solid residue of the tyre decomposition is transported by a conveying device installed inside the reactor, and is recovered at the tube end and dried. At a temperature of 500 "C, 54 wt% of oil, 26 wt% of carbon black, 11 wt% of gases and 9 wt% of steel are typically produced. Vacuum tyre pyrolysis allows the yield of the oil fraction to be increased with less production of gases and char compared with other tyre thermal decomposition methods. The oil produced was fractionated by distillation into three cuts: a naphtha fraction (bp < 204 "C,27 wt%), a light oil fraction (bp = 204-305 "C, 17 wt%) and a heavy oil fraction (bp > 350C, 56 wt%). The latter fraction was analysed and characterized in detail in order to consider its possible use as a feed in coking units. This heavy oil has a high specific gravity and is rich in polyaromatic hydrocarbons, which are the main precursors of coke. Moreover, compared with traditional coke feedstocks, the heavy oil fraction contains lower amounts of sulfur (0.8 wt%) and no vanadium, with a similar carbon content, although nitrogen compounds and several inorganic species are present in the pyrolytic oil. The heavy oil was subjected to delayed coking in a laboratory batch reactor operating under conditions similar to those used in industrial coking units (48O-50O0C, 1 atm), which led to the formation of the following fractions: coke (16.3 wt%), gases (6.1 wt%) and a liquid phase that again was separated by distillation into naphtha (13.5 wt%), light oil (45 wt%) and heavy oil (1 5.8 wt%). The authors proposed possible applications for the different products derived from the coking treatment. The gases formed are mainly methane and ethane, with low olefin content, and can be used as a fuel with high heating value. The naphtha fraction has a high aromatics content, hence it is suggested to be used as a gasoline component, although due to the presence of a certain amount of sulfur (0.3 wt%), nitrogen, olefins and diolefins, previous hydrotreatment of the naphtha will be required. The light oil fraction may be used as a fuel or further processed in fluidized bed catalytic units, either directly or after hydrogen upgrading to remove sulfur compounds (1.O wt% of sulfur). As far as the heavy oil produced during coking is concerned, it could be recycled to generate more coke. Finally, the coke obtained from the pyrolytic oil contains significantly lower amounts of sulfur and ash compared with standard petroleum coke. The main inorganic elements present in this coke are zinc and silicon. The absence of vanadium is noteworthy because this is an undesirable element in the composition of cokes. These results show the potential use of the tyre pyrolytic oil as a source of coke and other interesting fractions for refineries. Various alternatives have recently been proposed for the valorization of the solid carbon residue of tyre pyrolysis. Napoli et ~ 1 . have ' ~ ~compared the properties of the char with those of coal. The calorific value of the char is slightly higher than that of coal due to the lower ash content of the former, while

Therma 1 Processes

121

both materials have sulfur contents of about 1 wt%. Likewise, as shown in Figure 4.36, the particle size distribution of the char is narrower with a somewhat lower average size compared to coal. Accordingly, the authors conclude that co-combustion of coal and pyrolytic char, without any burner modification, is an interesting process provided the char content in the mixture is kept low due to the presence of zinc oxide and steel in this material. The feasibility of using the char generated by tyre pyrolysis as a precursor in the manufacture of activated carbon has been studied by various authors.' 19*13' Merchant and Petrich13' have obtained carbons with surface areas above 500 m2 g-' from tyre pyrolysis in batch reactors and subsequent activation of the chars by treatment with superheated steam at temperatures in the range 800-900 "C.Teng et al.' l 9 have obtained activated carbons with surface areas above 800 m2 g-' by pyrolysis of tyres up to 900 "C, followed by activation of the resulting chars in C02 at the same temperature. These surface areas are
6 100

5
- 4

60

E
$ . a 3

6o
40

g m
2
Y

22
1

20 0

0
10-1
(a)
6
100

10'

102

103

Parficle size (prn)

- 4

c
0
0)

4
60
40
20

$ 3 lY

c m
4-

2
1

0
10-1
(b)
100

10'

102

103

Particle size (prn)

Figure 4.36

Comparison of the particle size distribution o f a coal and a tyre pyrolysis char:'27(a) char, (b) coal. (Reprinted from J. Anal. Appl. Pyrol., 4 0 4 1 , A. Napoli, Y. Soudais, D. Lecomte and S. Castillo, page 373. 0 1997, with permission from Elsevier Science)

122

Chapter 4

comparable with those of commercial activated carbons. Moreover, these authors observed that the zinc and sulfur present in the final carbon residues are found spread throughout the char, but in the same locations, which suggests that they are present as zinc sulfide. In order to avoid char activation, which implies an additional process step and a decrease in the solid yield, San Miguel et al.'32 have explored the possible direct application of inactivated chars as adsorbents for the removal of pollutants in aqueous solutions. The chars were obtained by tyre decomposition in the temperature range 300-1000 "C, their textural properties being determined by nitrogen adsorption at 77 K. These data showed that the char pore structure is already formed at pyrolysis temperatures of about 600 "C and that it remains relatively unchanged above this temperature. The isotherms obtained were typical of mesoporous solids, although the presence of a certain microporosity was suggested by nitrogen adsorption at low relative pressures. BET surface areas between 75 and 85 m2 g- were obtained for chars prepared in the range 600-1000 "C. The adsorption properties of the chars were clearly inferior to those of commercial activated carbons for the removal from aqueous solutions of low molecular weight compounds, such as phenol, due to their lower surface area and microporosity. However, they showed adsorption capacities for two reactive dyes of large molecular size higher than those of the commercial activated carbon. Methods for the low temperature thermal decomposition of rubber wastes have also been reported, mainly in the presence of solvent or water. Ulick and C a r n e ~ -have ' ~ ~ developed a process for the degradation of vulcanized rubber based on treatment at 300C with an asphalt. After 1 h of reaction the starting tyre was completely degraded whereas the resulting products remained dissolved in the asphalt. Likewise, Saleh et a1.'34have reported that the treatment of butyl rubber pieces with water in a closed vessel at temperatures of about 350C caused the depolymerization of the raw material in 1 h. When starting from halobutyl rubbers, this process leads to a partial dehalogenation of the organic products. Similarly, the decomposition of rubber with supercritical water has been investigated by Funazukuri et af.135 Treatment of rubber with water at temperatures of about 380C led to a conversion of 4 3 4 8 % of the raw tyre into oil. Elementary analysis showed that about half of the sulfur contained in the original sample is extracted into the oil, whereas the other half remains in the solid residue.

'

7 Summary
Thermal degradation of plastic and rubber wastes in inert atmospheres has been extensively studied in the past. It is widely accepted that it takes place through radical mechanisms, two main pathways having been proposed: depolymerization by end-chain cracking and random chain scission. In the first case, high concentrations of the starting monomer are obtained, but this mechanism is predominant only in the thermal degradation of a few polymers, such as PS and

Thermal Processes

123

PMMA. For other polymers, random chain cleavage is the major pathway, which leads to the formation of very complex mixtures of products. The rate of the thermal decomposition of polyolefins is related to the presence of branches and side substituents in the polymer backbone. Accordingly, the following order of thermal degradation is usually observed: HDPE < LDPE < PP < PS. Depending on the temperature, up to four fractions can be produced by the thermal decomposition of these plastics: gases, oils, waxes and a solid residue. PVC degradation follows a different pathway: HC1 is first removed at a low temperature to yield a polyene residue, which is then decomposed at a higher temperature. During the processing of plastic mixtures and real plastic wastes, a crucial aspect is the existence of interactions between the different components, which can cause changes in the product distribution compared to the degradation of individual polymers. Thus, the HC1 released from PVC may interact with other degradation products to yield C1-containing organics or may act as a catalyst promoting the initial steps in the decomposition of other polymers. Likewise, the heat liberated during the reaction between HC1 and aluminium present in the wastes has been proposed to explain the sudden formation of waxy products, which is often observed in thermal processing plants for plastic wastes. A large number of processes and reactors have been developed for the thermal conversion of plastic and rubber wastes: stirred tanks, rotary kilns, fluidized beds, circulating bed reactors, screw extruders, etc. Many of the studies carried out in recent years have been based on sand fluidized or circulating bed reactors. Likewise, several works have recently appeared on plastic degradation in the presence of solvents. Of all the reaction conditions, temperature is the major variable determining the product distribution during the degradation of plastics. A number of high temperature (>600 "C) processes have been developed aimed at the production of large quantities of gases rich in olefinic hydrocarbons and to a lesser extent, aromatic oils. Enhanced production of ethylene and propylene has been obtained by working with systems with very high heating rates and short residence times. Likewise, the olefin yield can also be significantly increased by cracking the plastic in the presence of steam. On the other hand, several processes have recently been developed working at low temperature, usually around 500C. In this case, the main product is the oil fraction comprising mixtures of linear paraffins and a-olefins with little aromatic hydrocarbon content. It is proposed that these liquids are further upgraded by hydrogenation, preferentially in refinery units, although C1 contents below 10 ppm are required for this option. In other cases, the oils produced by the low temperature degradation of plastics have been proposed as a source of paraffins with a variety of potential applications. Finally, some studies have recently appeared on the possible thermal coprocessing of plastic wastes with coal and lignocellulosic materials, attempting to take advantage of the high hydrogen content of the former. Thermal conversion of rubber wastes, mainly used tyres, has also been widely investigated in the last two decades. In recent years, studies have focused on

124

Chapter 4

producing fractions of higher value with applications other than simply as low quality fuels. Typically, the thermal decomposition of used tyres leads to the formation of gases, oils and a solid residue or char. The gases formed are mainly H2, CO, C02, CH4 and C2H6, hence they can be used as a fuel gas. The oils are rich in aromatic hydrocarbons and have a lower sulfur content than the starting tyre. In most cases it is suggested that these oils are upgraded in refineries, although they may also be a source of chemicals. Finally, the chars produced, typically with yields of over 30 wt%, have a high calorific value but also a relatively high sulfur content. Different activation treatments have been described to increase the surface area of the chars in order to explore their possible use as activated carbons.

8 References
1 H.H.G. Jellinek, Degradation of Vinyl Polymers, Academic Press, New York, 1955. 2 S.L. Madorsky, Thermal Degradation of Organic Polymers, Interscience, New York, 1964. 3 N.S. Allen and M. Edge, Fundamentalsof Polymer Degradation and Stabilization, Elsevier Applied Science, New York, 1966. 4 L. Reich and S.S. Stivala, Elements of Polymer Degradation, McGraw-Hill, New York, 1971. 5 E.D. Owen, ed., Degradation and Stabilization of PVC, Elsevier, Amsterdam, 1984. 6 S.A. Liebman and E.J. Levy, eds., Pyrolysis and G C Polymer Analysis, Marcel Dekker, New York, 1985. 7 N. Grassie and G. Scott, Polymer Degradation and Stabilization, Cambridge University Press, London, 1985. 8 J. Wypich, Polyvinyl Chloride Stabilization, Elsevier, Amsterdam, 1986. 9 H.H.G. Jellinek, Degradation and Stabilization of Polymers, Elsevier, Amsterdam, 1988. 10 K.S. Minsker, S.V. Koselov, and G.E. Zaikov, Degradation and Stabilization of Vinylchloride-based Polymers, Pergamon Press, New York, 1988. 11 I. Ericsson, J . Anal. Appl. Pyrol., 1985,8, 73. 12 T.P. Wampler, J. Anal. Appl. Pyrol., 1989,15, 187. 13 L.A. Wall, J.H. Flynn, and S. Straus, Polym. Eng. Sci., 1970, 10, 19. 14 J.A. Conesa, A. Marcilla, R. Font, and J.A. Caballero, J. Anal. Appl. Pyrol., 1996, 36, 1. 15 S.S. Stivala, J. Kimura, and S.M. Gabbay, in Degradation and Stability of Polyolefins, ed. N.S. Allen, Applied Science, London, 1983. 16 R.R.A. Abou-Shaaban, J.L. Haberfled, E.M. Barrall, J.F. Johnson, and A.P. Simonnelli, Polym. Eng. Sci., 1976, 16, 544. 17 M. Seeger and R.J. Ritter, J . Polym. Sci., 1977, 15, 1393. 18 J.A. Conesa, A. Marcilla, and R. Font, J. Anal. Appl. Pyrol., 1994,30, 101. 19 R.W.J. Westerhout, J. Waanders, J.A.M. Kuipers, and W.P.M. van Swaaij, Ind. Eng. Chem. Res., 1998,37, 2316. 20 R.W.J. Westerhout, J. Waanders, J.A.M. Kuipers, and W.P.M. van Swajj, Ind. Eng. Chem. Res., 1997,36, 1955. 21 D.P. Serrano, R. Van Grieken, J. Aguado, J.C. Rojo, F. Temprano, R. Garcia, and J.A. Marugan (in preparation). 22 N. A. Sezgi, W.S. Cha, J.M. Smith, and B.J. McCoy, Ind. Eng. Chem. Res., 1998,37, 2582.

Thermal Processes

125

23 G.S. Darivakis, J.B. Howard, and W.A. Peters, Combust. Sci. Technol., 1990, 74, 267. 24 W. Kaminsky, J . Anal. Appl. Pyrol., 1985,8,439. 25 D.S. Scott, S.R. Czernik, J. Piskorz, and D.St.A.G. Radlein, Energy Fuels, 1990,4, 407. 26 J.A. Conesa, R. Font, and A. Marcilla, Energy Fuels, 1997, 11, 126. 27 E.A. Williams and P.T. Williams, J. Chem. Technol. Biotechnol., 1997,70, 9. 28 P.L. Beltrame, P. Carniti, G. Audisio, and F. Bertini, Polym. Degrad. Stab., 1989, 26, 209. 29 R.W.J. Westerhout, J. Waanders, J.A.M. Kuipers, and W.P.M. van Swajj, Ind. Eng. Chem. Res., 1998,37,2293. 30 S . Lovett, F. Berruti, and L.A. Behie, Ind. Eng. Chem. Res., 1997,36,4436. 31 V. Cozzani, C. Nicolella, M. Rovatti, and L. Tognotti, Ind. Eng. Chem. Res., 1997, 36, 324. 32 V. Cozzani, Ind. Eng. Chem. Res., 1997,36, 5090. 33 H. Sinn, W. Kaminsky, and J. Janning, Angew. Chem. Int. Ed. Engl., 1976, 15(1l), 660. 34 W. Kaminsky, B. Schlesselmann, and C. Simon, J. Anal. Appl. Pyrol., 1995,32, 19. 35 M. Gebauer, D. Schermaul, and D. Timm, Chem. Technik, 1995,47(4), 195. 36 B. Ondruschka, H.G. Struppe, J. Hofmann, U. Luther, J. Ahlheim, M. Gebauer, and D. Timm, Chem. Technik, 1995,47(4), 172. 37 M. Ratzsch, H. Grundmann, and R. Kilian, GB Patent A 1 108 295, 1968. 38 R.L. McConnell and D.A. Weemes, US Patent 3 519 609, 1970. 39 J. Dispons, Fr. Patent 2697 528, 1992. 40 M. Gebauer and J. Utzig, Chem. Technik, 1997,49(2), 57. 41 A. Chaala, H. Darmstadt, and C. Roy, J. Anal. Appl. Pyrol., 1997,39,79. 42 E. Kiran and J.K. Gillham, J. Appl. Polym. Sci., 1976, 20, 2045. 43 Y. Tsuchiya and K. Sumi, J . Polym. Sci., 1969, A-7, 1599. 44 J.K.Y. Kiang, P.C. Uden, and J.C.W. Chien, Polym. Degrad. Stab., 1980,2, 1 13. 45 H.H.G. Jellinek, J. Polym. Sci., 1948,3,850. 46 G.G. Cameron and J.R. McCallum, J. Macromol. Sci., Rev. Macromol. Chem., 1967, C1(2), 327. 47 G.G. Cameron and G.P. Kerr, Eur. Polym. J., 1968,4, 709. 48 I.C. McNeil, M. Zulfiqar, and T.A. Kousar, Polym. Degrad. Stab., 1990,28, 131. 49 A. Guyot, Polym. Degrad. Stab., 1986, 15,219. 50 H.H.G. Jellinek, J. Polym. Sci., 1948,4, 378. 51 B.V. Kokta, J.L. Valade, and W.N. Martin, J. Appl. Polym. Sci., 1973, 17, 1. 52 S.L. Malhotra, J. Hesse, and L.-P. Blanchard, Polymer, 1975, 16, 81. 53 K. Kishore, V.R. Verneker, and N.N.R. Nair, J. Appl. Polym. Sci., 1976,20, 2355. 54 T. Kuroki, T. Ikemura, T. Ogawa, and Y. Sekiguchi, Polymer, 1982,23, 1091. 55 A. Marcilla and M. Beltran, Polym. Degrad. Stab., 1995,50, 117. 56 G . Madras, G.Y. Chung, J.M. Smith, and B.J. McCoy, Ind. Eng. Chem. Res., 1997, 36, 2019. 57 P. Carniti, A. Gervasini, and P.L. Beltrame, J. Polym. Sci., Part A: Polym. Chem., 1989,27,3865. 58 U.K.O. Schroder and K.H. Ebert, Makromol. Chem., 1984,185,991. 59 Z. Zhang, T. Hirose, S. Nishio, Y. Morioka, N. Azuma, and A. Ueno, Ind. Eng. Chem. Res., 1995,34,4514. 60 P. Carniti, P.L. Beltrame, M. Armada, A. Gervasini, and G. Audisio, Ind. Eng. Chem. Res., 1991,30, 1624. 61 P.T. Williams, P.A. Horne, and D.T. Taylor, J. Anal. Appl. Pyrol., 1993, 25, 325. 62 T. Murakata, Y. Saito, T. Yosikawa, T. Suzuki, and S. Sato, Polymer, 1993,34(7), 1436. 63 S . Sato, T. Murakata, S. Baba, Y. Saito, and S. Watanabe, J . Appl. Polym. Sci., 1990,40,2065.

126

Chapter 4

64 T. Murakata, S. Wagatsuma, Y. Saito, T. Suzuki, and S. Sato, Polymer, 1993,34(7), 143 1. 65 G. Madras, J.M. Smith, and B. McCoy, Ind. Eng. Chem. Res., 1995,34,4222. 66 P.L. Beltrame, L. Bergamasco, P. Carniti, A. Castelli, F. Bertini, and G. Audisio, J . Anal. Appl. Pyrol., 1997,4041,451. 67 G. Audisio and F. Bertini, J . Anal. Appl. Pyrol., 1992,24,61. 68 B. Baum and L.H. Wartman, J. Polym. Sci.,1958,28,537. 69 R.R. Stromberg, S. Straus, and B.G. Achhammer, J. Polym. Sci., 1959,35, 355. 70 J.D. Danforth and T. Takeuchi, J . Polym. Sci., Part A: Polym. Chem., 1973, 1 1 , 209 1. 71 E.P. Chang and R. Salovey, J . Polym. Sci., Part A: Polyrn. Chem., 1974,12,2927. 72 M.P. Farr and I.R. Harrison, J. Polym. Sci., Part C: Polym. Lett., 1986, 24, 257. 73 Z. He, X. Hu, and G. Sun, Polym. Degrad. Stab., 1989,24, 127. 74 D. Braun, Pure Appl. Chem., 1971,26, 173. 75 G . Ayrey, B.C. Head, and R.C. Poller, J. Polym. Sci., Macromol. Rev., 1974, 8, 1. 76 D. Braun, B. Bohringer, B. Ivan, T. Kelen, and F. Tudos, Eur. Polym. J., 1986, 22, 229. 77 J. Behnish and H. Zimmermann, J. Thermal Anal., 1988,33, 191. 78 A.J. Chirinos-Padron and G.A. von Schoettler, Polym. Degrad. Stab., 1991,33,213. 79 B.B. Troitskii and L.L. Troitskaya, J. Polym. Sci., Part A: Polyrn. Chem., 1993,31, 75.

80 C.-H. Wu, C.-Y. Chang, J.-L. Hor, S.-M. Shih, L.-W. Chen, and F.-W. Chang, Can. J . Chem. Eng., 1994,72, 644. 81 A. Jimenez, V. Berenguer, J. Lopez, and A. Sanchez, J. Appl. Polym. Sci., 1993,50,
1565. 82 T. Hjertberg and E.M. Sorvik, J . Appl. Polym. Sci., 1978,22,2415. 83 E. Martisson, T. Hjertberg, and E.M. Sorvik, Macromolecules, 1988, 21, 136. 84 K. Patel, A. Velazquez, H.S. Calderon, and G.R. Brown, J. Appl. Polym. Sci.,1992, 46, 179.

85 A.B.J. Oudhuis, P. De Wit, P.J.J. Tromp, and J.A. Moulijn, J. Anal. Appl. Pyrol.,
1991, 20, 321. 86 87 88 89 90 91 92 93

A. Marcilla and M. Beltran, Polym. Degrad. Stab., 1996,53,261. A. Marcilla and M. Beltran, Polym. Degrad. Stab., 1996,53,251. M.C. Gupta and S.G. Viswanath, Ind. Eng. Chem. Res., 1998,37,2707. M. Blazso, J. Anal. Appl. Pyrol., 1997,4041,69. W. Kaminsky and J. Franck, J . Anal. Appl. Pyrol., 1991,19, 311. G. Madras, J.M. Smith, and B.J. McCoy, Ind. Eng. Chem. Res., 1996,35, 1795. W. Kaminsky, Makromol. Chem., Macromol. Symp., 1991,48-49,381. C. Shimasaki, N. Watanabe, K. Fukushima, S. Rengakuji, Y. Nakamura, S. Ono, T. Yoshimura, H. Morita, M. Takakura, and A. Shiroishi, Polym. Degrad. Stab.,

1997,58, 171. 94 C.-H. Wu, C.-Y. Chang, J.-L. Hor, S.-M. Shih, L.-W. Chen, and F.-W. Chang, Waste Management, 1993, 13,221. 95 Y. Sakata, M.A. Uddin, K. Koizumi, and K. Murata, Polym. Degrad. Stab., 1996, 53, 111. 96 M. Blazso, B. Zelei, and E. Jakab, J . Anal. Appl. Pyrol., 1995,35,221. 97 Y. Sakata, M.A. Uddin, A. Muto, M. Narazaki, K. Koizumi, K. Murata, and M. Kaji, Ind. Eng. Chem. Res., 1998, 37, 2889. 98 R. Rausa and P. Pollesel, J . Anal. Appl. Pyrol., 1997,4041,383. 99 J . Kim, W. Kaminsky, and B. Schlesselmann, J . Anal. Appl. Pyrol., 1997, 40-41, 365. 100 W. Kaminsky and H. Rossler, Chemtech, 1992,2, 108. 101 W. Kaminsky, B. Schlesselmann, and C.M. Simon, Polym. Degrad. Stab., 1996,53, 189. 102 M.A. Paisley and R.D. Litt, US Patent 5 326919, 1994.

Thermal Processes

127

E.A. Williams and P.T. Williams, J. Anal. Appl. Pyrol., 1997,4041, 347. Oil Gas J., 1997,95(33), 18. H. Kastner and W. Kaminsky, Hydrocarbon Processing, 1995,74(5), 109. K.C. Kirkwood, S.A. Leng, and D.W. Sims, US Patent 5 364 995, 1994. H.P. Wenning, J. Anal. Appl. Pyrol., 1993,25,301. J. Utzig and M. Gebauer, Proceedings of the R97 Congress, Geneva, 1997. H. Coenen and R. Hagen, US Patent 4642401,1987. Y. Tsukagoshi, K. Tsuchiya, T. Inoue, T. Ogawa, H. Komori, K. Murakami, and T. Ayukawa, US Patent 3 947 256, 1976. 11 1 W. Michaeli and V. Lackner, Angew. Makromol. Chem., 1995,232, 167. 112 M.W. Meszaros, Proceedings of Recyclingplas VI 91, Washington, 1991, p. 335. 113 U. Hofmann and M. Gebauer, Kunststofle German Pfast., 1993,83(4), 8. 114 S.R. Palmer, E.J. Hippo, D. Tandon, and M. Blankenship, in Coal Science, ed. J.A. Pajares and J.M.D. Tascon, Elsevier Science, Amsterdam, 1995, p. 1523. 115 K. Miura, K. Mae, T. Maki, and M. Takase, in Coal Science, ed. J.A. Pajares and J.M.D. Tascon, Elsevier Science, Amsterdam, 1995, p. 1527. 116 G . Collin and J. Polaczek, in Coal Science, ed. J.A. Pajares and J.M.D. Tascon, Elsevier Science, Amsterdam, 1995, p. 1033. 117 P. Sivakumar, H. Jung, J.W. Tierney, and I. Wender, Fuel Processing Technol., 1996,49, 219. 118 C. Di Blasi, J. Anal. Appl. Pyrol., 1997,4041,463. 119 H. Teng, M.A. Serio, M.A. Wbjtowicz, R. Bassilakis, and P.R. Solomon, Ind. Eng. Chem. Res., 1995, 34, 3102. 120 P. Maruhnic, R.H. Wolk, and C.A. Battista, US Patent 3 978 199, 1976. 121 B.L. Shulman and P.A. White, in Solid Wastes and Residues: Conversion by Advanced Thermal Processes, ed. J.L. Jones and S.B. Radding, ACS Symposium Series 76, American Chemical Society, Washington DC, 1978, p. 274. 122 H. Pakdel, C. Roy, H. Aubin, G. Jean, and S. Coulombe, Environ. Sci. Technol., 1991,25(9), 1646. 123 P.T. Williams, S. Bestler, and D.T. Taylor, Proc. Inst. Mech. Eng., 1993,207, 55. 124 B. Benallal, C. Roy, H. Pakdel, S. Chabot, and M.A. Poirier, Fuel, 1995, 74(1 l), 1589. 125 P.T. Williams, S. Besler, D.T. Taylor, and R.P. Botrill, J . Inst. Energy, 1995,68, 11. 126 P.T. Williams, S. Besler, and D.T. Taylor, Fuel, 1990, 69, 1474. 127 A. Napoli, Y. Soudais, D. Lecomte, and S. Castillo, J. Anal. Appl. Pyrol., 1997,4& 41, 373. 128 C. Roy, US Patent 4 740 270, 1988. 129 A. Chaala and C. Roy, Fuel Processing Technol., 1996,46,227. 130 J.A. Conesa, R. Font, and A. Marcilla, Energy Fuels, 1996,10, 134. 131 A.A. Merchant and M.A. Petrich, AICHE J., 1993,39, 1370. 132 G. San Miguel, G.D. Fowler, and C.J. Sollars, Ind. Eng. Chem. Res., 1998,37,2430. 133 T.J. Ulick and W.E. Carner, US Patent 5070 109, 1991. 134 R.Y. Saleh, M. Siskin, and G. Knudsen, US Patent 5 283 318, 1994. 135 T. Funazukuri, T. Takanashi, and N. Wakao, J. Chem. Eng. Jpn., 1987,20(1), 23. 103 104 105 106 107 108 109 110

CHAPTER 5

Catalytic Cracking and Reforming


1 Introduction
The processes of feedstock recycling of plastic wastes considered in this chapter are based on contact of the polymer with a catalyst which promotes its cleavage. In fact, plastic degradation proceeds in most cases by a combination of catalytic and thermal effects which cannot be isolated. As was described in Chapter 3, the use of catalysts is also usual in chemolysis processes of plastic depolymerization. However, there are two main differences between catalytic cracking and chemolysis: there is no chemical agent incorporated to react directly with the polymer in catalytic cracking methods, and the products derived from the polymer decomposition are not usually the starting monomers. Compared to the simple cleavage of the polymer by thermal effects, catalytic cracking has a number of advantages:
0

The polymer molecules start to break down in the presence of catalysts at considerably lower temperatures than in thermal decomposition. A significant catalytic conversion of polyolefins into volatile products has been detected at temperatures as low as 200C, compared with the value of 400 "C which is necessary in the thermal degradation of PE and PP to observe the formation of the first gases. As a consequence, catalytic treatments of plastic materials are usually carried out at low temperatures, in contrast with the range of 500-800C, typical for thermal cracking and pyrolysis. When compared at the same temperature, catalytic cracking of polymers proceeds faster than thermal degradation, i.e. with lower activation energy. At temperatures of about 400"C, the first volatile products are observed after only a few minutes of contact with the catalyst. The products derived from the catalytic cracking of plastics are of higher quality than those obtained by thermal decomposition. Thus, the presence of a high proportion of branched, cyclic and aromatic structures in the oils produced lead to properties very similar to those of commercial
129

130

Chapter 5

gasolines. Moreover, the product distribution can be varied and controlled by the selection of a suitable catalyst and modification of its properties. All these factors point out the great potential of catalytic cracking for the conversion of polymeric wastes into valuable products. However, this method also suffers from a number of drawbacks and problems, which are still not completely solved. The catalysts are deactivated with time by the deposition of carbonaceous residues and poisons present in the raw waste stream, such as C1 and N compounds. Moreover, the inorganic compounds contained in the plastic wastes tend to remain with the catalysts, hindering their recovery and reuse. For these reasons, catalytic cracking is mainly applied to polyolefinic wastes of relatively high purity, a number of pretreatments being required to remove all those components which may negatively affect the catalyst. Other difficulties arise from the high viscosity of the molten plastic, which hinders its flow through conventional fixed bed reactors. These problems are largely avoided when the catalytic conversion is combined with a simple thermal treatment, aimed at reducing the viscosity of the mixture and enabling the separation of unwanted components. In this case, the catalytic step consists really of a reforming of the products directly formed by thermal degradation of the polymers. In the following sections of this chapter, the catalytic conversion of individual plastics (polyethylene, polypropylene and polystyrene) is first reviewed, followed by a description of the processes developed for the catalytic cracking of plastic and rubber mixtures. Finally, methods based on a combination of thermal and catalytic treatments are considered. However, taking into account that the key factor in the catalytic conversion of plastic wastes is the catalyst itself, we will first describe the main properties of the most widely used catalytic systems for the degradation of polymers.

2 Types and Properties of Polymer Cracking Catalysts


A wide variety of catalysts have been found effective in promoting the decomposition of plastic materials: Friedel-Crafts catalysts, acidic and basic solids, bifunctional solids, etc. Friedel-Crafts systems, mainly AlC13/HCl, were initially used as acid catalysts but they have now been replaced in most processes by solids with acid properties due to the corrosion and environmental problems they cause. The most common catalysts used in plastic cracking are acidic solids, mainly alumina, amorphous silica-alumina and zeolites. These materials are the catalysts typically used in the petroleum processing and petrochemical industries. They have very different textural and acid properties, which directly determine their catalytic activity and product selectivity. Thus, while the acidity of alumina is of Lewis type, both Bronsted and Lewis acid sites may be present in amorphous silica-alumina and zeolites. This is an important factor because

Catalytic Cracking and Reforming

131

the initiation step of polymer catalytic degradation depends on the type of acid site: by proton addition over Bronsted sites and by hydride abstraction over Lewis sites, which leads to different cracking pathways, The concentration of acid sites in both amorphous silica-alumina and zeolites can be controlled by changing the Si/A1 ratio, because the acid sites are generated by A1 species. Moreover, the A1 content of the catalyst usually influences the strength of the acid sites. Stronger sites favour the cracking reactions, although they may also promote undesired reactions, such as coke deposition which can cause catalyst deactivation. Alumina and amorphous silica-alumina are usually mesoporous materials with a wide distribution of pore sizes. The surface area, pore size and pore volume of alumina and amorphous silica-alumina depend greatly on the preparation method, hence their textural properties can be controlled to a certain extent by changing the synthesis conditions. These parameters are also highly relevant in determining the catalytic properties of these materials. On the contrary, zeolites are by definition microporous crystalline silicoaluminates. They have a perfectly defined crystalline structure based on the linkage between SiO4 and A104 tetrahedra through oxygen bridges. The presence of pores with sizes below 1.O nm in zeolitic structures allows different molecules to enter, diffuse and react within them. At present there are over 100 known zeolite structure types,' some of which occur naturally, although those having the most significant catalytic applications are synthetic materials. Zeolites are classified according to their pore size (small, medium and large pore zeolites), the number of channel systems (unidimensional, bidimensional and tridimensional pore zeolites), and A1 content (low, medium and high silicalalumina ratio zeolites). Depending on the topology and the preparation method, the A1 content of zeolites can be varied over a wide range, from a Si/Al ratio of unity to negligible amounts of aluminium.* The acid form of zeolites is obtained when the negative charge associated with the framework A1 species is balanced by protons. The A1 content can be varied by synthesis or post-synthesis methods and it is usually accompanied by significant changes in the zeolite acid strength. Moreover, both Bronsted and Lewis acid sites can be present in zeolites, depending on their structure and the Si/A1 ratio. Protons can be ion exchanged by other metal cations, which affects both the acid properties and the effective pore size. Moreover, zeolites with basic catalytic properties can be generated by ion exchange with cations such as Cs. The features of the zeolite channel systems are also key factors in explaining their catalytic properties. The possibility of discriminating between reactants and products according to their molecular size compared with the zeolite pore diameter is a widely reported phenomena, known as shape ~electivity.~ Compared with amorphous silica-alumina, zeolitic catalysts exhibit a number of advantages: a higher acid strength related to their crystalline structure, narrower distribution of pore sizes, higher stability under thermal and hydrothermal conditions, lower rate of coke deposition, etc. The most common zeolites are also those which have been most extensively used for the catalytic cracking of polymers: X, Y , ZSM-5, mordenite, etc. Figure 5.1 shows the

132

Chapter 5

A) ZEOLITE ZSM-5

B) ZEOLITE Y

Figure 5.1 Pore structure of dzferent zeolites: (A) ZSM-5, (B) Y.

structure of zeolites Y and ZSM-5, and the main features of several zeolites are summarized in Table 5.1. As a consequence of all these properties, zeolites can be considered to be exceptional catalysts which have replaced amorphous solids in many applications. However, for the catalytic cracking of polymeric wastes, zeolites may be disadvantageous due to the steric and diffusional problems that polymer molecules may have in accessing the zeolite micropores. These drawbacks can be overcome with the use of zeolitic catalysts with very small crystal size and, therefore, with a high proportion of external surface area which is not subjected to steric hindrances for the conversion of bulky substrates. Other interesting solids for the catalytic degradation of polymeric wastes are the various silica-based mesophases which have recently been d i s ~ o v e r e d . ~ ~ These materials are characterized by the presence of ordered and regular pore systems and high surface areas, typically over 1000 m2 g - . The most common member of this family is MCM-41, which has a hexagonal array of uniform pores with diameters that can be tailored in the range 1.5-10 nm by varying the synthesis conditions. These mesoporous materials can be prepared with a wide

Catalytic Cracking and Reforming

133

Table 5.1 Main properties of various zeolites


Zeolite Structure SiIAI Ratio Pore size (nm)

ZSM-5 Y Beta Mordenite

M FI FAU BEA MOR

1&1000 1.5-3* 8-1 000


5*

0.53 x 0.56,0.51 x 0.55 0.74 0.76 x 0.64,0.55 x 0.55 0.65 x 0.7

*The Si/AI ratio of zeolites Y and mordenite can be increased by post-synthesis treatments.

range of framework compositions and exhibit properties of both conventional zeolites and amorphous silica-alumina: uniformity of pore sizes, regular pore ordering, amorphous nature of the pore walls, presence of both Lewis and Bronsted acid sites of weak and medium strength, etc. The combination of high surface area with uniform mesopores in MCM-41 are the reasons which support the remarkable catalytic properties that this material has recently shown for the conversion of po~yolefins.~.~ In addition to silica- and alumina-based solids, activated carbon impregnated with transition metals and sulfated zirconia have also been tested in the catalytic cracking of organic polymers. Activated carbons are microporous solids with a graphitic-like structure and large surface areas. The incorporation of transition metals on activated carbon leads to the generation of hydrogenating/dehydrogenating active sites, which promotes hydrogen transfer reactions during plastic decomposition. Sulfated zirconia is known as a superacid solid, i.e. its acid strength is greater than that of 100% H2SO4. Thus, the Hammet acidity of sulfated zirconia is - 16.04, appreciably higher than the value of - 11.94 corresponding to 100% H2SO4. Sulfated zirconia is therefore commonly used as a catalyst in reactions requiring strong acid sites.

3 Catalytic Conversion of Individual Plastics


Polyethylene
Most of the studies reported on the catalytic cracking of plastics use PE as starting material because it is the main polymer in plastic wastes. The first works appeared in the 1 9 7 0 ~ mostly ,~ based on the use of Friedel-Crafts catalysts. Ivanova et al.9 have described in detail the mechanism of PE degradation over A1C13-based catalysts. Compared with thermal decomposition, the catalytic conversion of PE at 400 "C leads to higher conversions with significant changes in the product distribution, because the catalyst also promotes secondary isomerization reactions. While the thermal decomposition of PE takes place via a radical mechanism, the catalytic cracking over acid solids proceeds through carbocationic species. According to Ivanova et a ~the, initiation ~ step involves the formation of carbocations through hydride abstraction from the polymeric chains by the H +[A1Cl3-OH-] catalyst with release of H2. This attack

134

Chapter 5

may take place at random positions in the polymer chains or, if side branches are present, selectively at tertiary carbons. A preferential degradation may also occur at double bond defects, in this case by proton addition. The carbocations formed are isomerized to yield more stable tertiary cationic species, which in turn undergo p-scission or elimination of the side chain. Other reactions which occur in the system are intermolecular and intramolecular transfer processes, similar to radical mechanisms. Beltrame et al. l o have compared the thermal degradation of PE with catalytic cracking over amorphous silica-alumina and zeolite catalysts. Figure 5.2 shows the TGA corresponding to the PE degradation alone and in the presence of 10 wt% of zeolite HY (protonic form of zeolite Y). While thermal conversion of PE into volatile compounds starts at 400"C, the catalytic decomposition causes a significant weight loss at temperatures below 200 "C and the polymer is almost completely converted before reaching 400 "C. The authors have also determined the activation energy of the PE degradation from TGA data, a significant reduction being observed in the presence of the catalysts. In contrast with the high activation energy found for the thermal degradation (273 kJ mol-'), catalytic degradation over zeolites HY and REY (rare earth exchanged form of zeolite Y) led to values of 57.3 and 41.0 kJmol-'. A significant decrease was also observed for silica-alumina materials, with activation energies around 120 kJ mol - This effect was less significant when silica or alumina were used as catalysts, activation energies of 193.1 and 181.8 kJ mol-', respectively, being calculated. According to these results, the decrease in the activation energy of PE degradation appears to be correlated with the acid strength of the catalysts.

'.

Figure 5.2

TG analysis in nitrogen of PE alone and a P E + H Y mixture (10 wt% o f HY)." (Reprinted from Polym. Degrad. Stab., 26, P.L. Beltrame, P. Carniti, G . Audisio and F. Bertini, page 209. 0 1989, with permission from Elsevier Science)

Catalytic Cracking and Reforming

135

The presence of these catalysts also leads to meaningful variations in the product distribution. Figure 5.3 compares the GC analysis of the oil fraction obtained by Mordi et a1.l' in the degradation of LDPE alone and in the presence of a ZSM-5 catalyst at 350C. The products of the thermal degradation were distributed over a wide range of carbon atom numbers, the peaks appearing as doublets or even triplets, which indicates the presence of the corresponding linear alkane, alkene and diene for each atom carbon number. In contrast, the and several peaks are present products of the catalytic oil are mainly below C15, for each atom carbon number, indicating the presence of multiple isomers, mainly alkanes and aromatics, whereas the amount of a-olefins is negligible. The mechanism and products of the PE catalytic decomposition over amorphous silica-alumina catalysts has been investigated by Ishihara et al.,12?13 using a batch reactor under nitrogen atmosphere. Four fractions were recovered after the reaction: gases, liquids, degraded oligomer and degraded polymer. The last two fractions were obtained from the residue remaining in the reactor by partial dissolution in xylene, the degraded polymer being the solubilized fraction. The formation of gases and oils was detected at temperatures as low as 220"C, although operating with high catalyst concentration (catalyst/polymer mass ratio, C/P = 1). The authors propose a sequential mechanism to explain the catalytic PE degradation: polymer -+degraded polymer -+ degraded oligomer + liquids + gases

12 11

fo.

14

k
Figure 5.3 components identified: o-xylene ( 8 a ) , p + m-xylene ( 8 b ) , naphthalene (1Oa), n-methylbenzene ( 1Ob) , methyl-naphthalene ( 1l a ) , solvent ( S ). (Reprinted from J . Anal. Appl. Pyrol., 29, R.C. Mordi, R. Fields and J. Dwyer, page 45. 0 1994, with permission from Elsevier Science)
GC analysis of the oils derived f r o m LDPE degradation at 350 OC:" (A) thermal conversion, (B) catalytic conversion over ZSM-5. Aromatic

136

Chapter 5

The gases produced contained over 50mol% of isobutane, and minor amounts of isobutene, propylene, propane, isopentane, etc. Interestingly, the liquid degraded oligomer and degraded polymer fractions were free of olefinic compounds, in contrast to the high concentration of double bonds typically present in the products of thermal degradation. The 13C-NMR spectrum of the degraded oligomer fraction, shown in Figure 5.4, indicates that these oligomers contain a large number of short branches, most of them C6 or less in length. The major types include short linear chains, ranging from methyl to pentyl, and some branched alkyl chains. The authors found a relationship between the molecular weight reduction and the branching frequency of degraded oligomers and polymer. Figure 5.5 shows the concentration of branched methyl groups, excluding those at the end of the backbone, versus the average molecular weight. It is concluded that the number of side chains increases linearly with the reduction in the molecular weight of the degraded polymer. The high degree of branching is due to the isomerization activity of the amorphous silica-alumina catalyst, which promotes conversion of secondary carbocations into tertiary ones, rather than p-scission. It is proposed that gases are formed selectively from the liquid components having the highest branching frequency. Ishihara et d i 4 have also studied the catalytic degradation of PE over amorphous silica-alumina in a continuous flow fixed bed reactor. Careful measurements of the temperature gradients in the reactor showed that some of the PE degradation takes place by thermal cracking in the high temperature zones located prior to the catalytic fixed bed. Therefore, when the molten

40
Figure 5.4

30

20

10

ppm from TMS


I3C-NMR spectrum of the oligomer fraction obtained by PE catalytic degrada t ion o ver amo rplio us silica-alum ina . (Reprinted from J . Appl. Pol-vm. Sci., 38, Y. Ishihara, H. Nambu, T. Ikemura and T. Takesue, page 1491. 0 1989, with permission from Elsevier Science)

Catalytic:

I37

Figure 5.5

Relationship between the molecular weight reduction and the branching concentration during the P E degradation over amorphous silica-alumina. (Reprinted from J. Appl. Polym. Sci., 38, Y. Ishihara, H. Nambu, T. Ikemura and T. Takesue, page 1491. 0 1989, with permission from Elsevier Science)

'

polymer is contacted with the catalyst, it is really made up of thermally degraded oligomers. Under these conditions, very high yields of gaseous products were observed with an overall polymer conversion of about go%, in spite of the low residence time in the catalyst bed ( - 20 s). Gas yields of over 70 wt% were obtained in the temperature range 400475 "C. In this system, the previous thermal cracking contributes not only to a decrease in the molecular weight, but also to a generation of olefin groups in the oligomers, which favours the subsequent catalytic decomposition by proton addition from the catalysts to both internal and chain end double bonds. The effect of PE properties on its catalytic degradation over amorphous silica-alumina catalysts has recently been studied by Uddin et af.I5in a batch reactor at 430C. Four different types of PE were used: high density PE (HDPE), low density PE (LDPE), linear low density PE (LLDPE), and crosslinked PE (XLPE). Compared with thermal degradation, in all cases the catalytic conversion was 3-4 times faster. In both thermal and catalytic degradation, HDPE and XLPE were converted at a slower rate, indicating the greater difficulty of cracking these polymers, probably as a consequence of the lower concentration of branches in their polymer chains. In the catalytic experiments, conversions of about 90% were obtained and no waxy products were formed. The main products were liquids with yields of about 80%. Only slight changes were detected in the product distribution corresponding to the catalytic cracking of the four PE types. PE catalytic degradation over zeolite NaY was also investigated by Ishihara et af.16in a batch reactor. The authors propose a sequential mechanism similar

138

Chapter 5

to that previously described over amorphous silica-alumina. One of the major differences found is the higher proportion of isopentane present in the gases produced over the zeolite. The oligomer fraction was composed mainly of saturated structures with a high number of methyl branches, and some olefinic signals were detected by 'H-NMR in the liquid fraction. It is proposed that isobutane and isopentane are generated by decomposition of Cg species present in the liquid fraction. Lin et al.17 have investigated, using TG measurements, the deactivation of the zeolite USY (ultrastable zeolite Y) by coke deposition during the degradation of HDPE. The cracking activity of previously coked zeolite samples was determined by successive reactions with fresh polymer. A significant decrease in activity was observed, which led to a progressive shift of the HDPE degradation towards higher temperatures. The activity was found to decline exponentially with the coke content. Thus, half of the initial catalytic activity is lost for a coke content of 4.24 wt% in the catalyst. Nevertheless, most of the initial activity can be recovered by calcination of the catalyst in air, as can be seen in Figure 5.6 showing the T G analysis of HDPE and USY mixtures through several reactionregeneration cycles. The catalytic properties, including resistance to deactivation, of different zeolites (HY, H-mordenite and ZSM-5) and amorphous silica-alumina have recently been compared by Uemichi et a1.18 in a fixed bed reactor. The yields of the different products obtained at 450C are illustrated in Figure 5.7, whereas
100

90

10

0
150
200 250

300

350

400

450

500

Temperature /"C
Figure 5.6

TG anal sis of HDPE cycles. ,I'

+ USY mixtures through several reaction-regeneration

(Reprinted from Thermochim. Acta, 294, Y.-H. Lin, P.N. Sharratt, A.A. Garforth and J. Dwyer, page 45. 0 1997, with permission from Elsevier Science)

Catalytic Cracking and Reforming

139
oLiquid
EWax

80
mGas
60
n

.Coke

40

9
Q)

20

ZSM-5
Figure 5.7

HY

HM

SA

Yields of products obtained in the catalytic degradation of PE at 450 "C in a jixed bed reactor over diferent catalysts:'8 Z S M - 5 , H Y , H-mordenite ( H M ) and amorphous silica-alumina ( S A ) . (Reprinted with permission from Y. Uemichi, M. Hattori, T. Itoh, J. Nakamura and M. Sugioka, Ind. Eng. Chem. Res., 37,867. 0 1998 ACS)

the variation in coke deposition with time on stream is shown in Figure 5.8. ZSM-5 was very active for the conversion of PE, mainly into gaseous products. A high initial activity was also observed over the HY zeolite, but a large amount of coke is formed, which leads to a very rapid deactivation of this catalyst with increasing time on stream. The fast deactivation of HY is confirmed by the coke content of about 20 wt% detected in this sample. The presence of large cavities in the structure of zeolite Y favours the formation of polyaromatic species, which are coke precursors. In contrast, the main products derived from PE degradation over amorphous silica-alumina and H-mordenite are liquid hydrocarbons, although a high proportion of wax was also produced on the latter, showing the lower activity of this zeolite. Moreover, H-mordenite was very rapidly deactivated, probably due to the unidimensionality of its channel system, which is easily blocked by coke deposition. The best catalyst was ZSM-5 zeolite, in terms of both activity and deactivation. The liquid yield over ZSM-5 remained constant, and coke deposition on this zeolite levelled off after 150 min. Likewise, amorphous silica-alumina resulted in a stable catalyst in spite of the rapid coke deposition which occurs. In this case, the large pores of the amorphous silica-alumina (2-8 nm) means that this material is little affected by coke deposition, at least in the range of time on stream investigated. The liquid produced over the ZSM-5 zeolite has a boiling point in the range of commercial gasolines. GC-RON (gas chromatography-research octane number) measurements indicate that this liquid can be considered a high quality gasoline with research octane numbers between 103 and 112, as a

140

Chapter 5

> a

in

Oe20

3 0.15

eZSM-5 .SA 0HY OHM

Time on stream (min)


Figure 5.8 Evolution of coke formation with time on stream during the catalytic degradation of P E at 450 "C in aJixed bed reactor over diferent catalysts:'8 ZSM-5, H Y , H-mordenite ( H M ) and amorphous silica-alumina ( S A ) . (Reprinted with permission from Y. Uemichi, M. Hattori, T. Itoh, J. Nakamura and M. Sugioka, Ind. Eng. Chem. Res., 3 7 , 8 6 7 . 0 1998 ACS)

consequence of its high aromatic content. In contrast, the liquid obtained over the amorphous silica-alumina sample had a lower RON number (about 85). Lin and White" have also compared the activity of different acid catalysts (amorphous silica-alumina, ZSM-5 zeolite and sulfated zirconia) for PE degradation based on thermogravimetry-mass spectrometry (TG-MS) measurements. The rate of PE catalytic cracking and the reduction observed in the activation energy were directly related to the acid strength of the catalyst, being in the following order: sulfated zirconia > ZSM-5 > amorphous silicaalumina. For the three catalysts investigated, PE degradation proceeds in two stages at different temperatures. The first maximum in the formation rate of volatile products may result from catalyst protons attacking defective double bonds of the polymer, whereas the second maximum at higher temperature may reflect the acid-catalysed cracking of the -CH2- polymer backbone. The main components of the gases produced were olefins (propylene and isobutene), which is in contrast with the high proportion of isoalkanes found in previous studies of PE catalytic conversion. The authors propose that olefins are transformed into alkanes by protonation of the double bond followed by hydrogen abstraction from the polymer, but these reactions do not take place in TGA tests, probably due to the rapid removal of the volatile products from the reaction medium. ZSM-5 zeolite produced the largest fraction of aromatics, which was related to the restricted channel volume of this zeolite which favours oligomerization reactions of olefins to form small alkyl aromatics. This conclu-

Catalytic Cracking and Reforming

141

sion is supported by the fact that aromatics are detected only after the formation of significant quantities of alkenes. In a recent work the catalytic degradation of both HDPE and LDPE was investigated by Aguado et al.7 in a batch reactor at 400 "C over three different catalysts: the mesoporous material MCM-4 1, zeolite ZSM-5, and an amorphous silica-alumina sample. For both polymers, the activity order found was as follows: ZSM-5 > MCM-41 > > amorphous silica-alumina. The higher conversions obtained over the zeolite were assigned to its stronger acidity, whereas the superior activity of MCM-41 compared to the silica-alumina was related to the large surface area of MCM-41 (over 1000 m2 g-'). Similarly to the thermal cracking reactivity of these two polymers, the catalytic cracking of LDPE was faster than HDPE over the three catalysts investigated, probably due to the higher degree of branching in its backbone. Figure 5.9 compares the product distribution obtained in the LDPE and HDPE degradation over MCM-41 and ZSM-5. Polyolefin cracking over the zeolite leads to a high proportion of gaseous hydrocarbons rich in olefins and a liquid fraction in the range of gasolines (C5-C12) with a high aromatic content. Over MCM-41 less olefinic gases are generated, while, in addition to a gasoline fraction, middle distillates with hydrocarbons in the range CI3-C22 are also produced. This product distribution is in agreement with the pore size of the catalysts, middle distillates only being formed over the mesoporous material. Comparing the two types of PE, it is seen that with both catalysts higher amounts of oils are obtained in the LDPE cracking. Sakata et a1.20have also studied polyethylene degradation over a mesoporous silica catalyst. The material used, called KFS- 16, is closely related to MCM-41, although it is prepared by a different method starting from a layered silicate (kanemite). One of the most interesting observations in this work is the PE cracking activity exhibited by KFS-16 in spite of the absence of acid sites (it is a completely silica-based material). Figure 5.10 shows the cumulative volume of liquid products obtained in the thermal and catalytic cracking of PE in a batch

Figure 5.9

Product distribution obtained in the catalytic conversion of HDPE and L D P E at 400 "C in a batch reactor over Z S M - 5 zeolite and MCM-41 catalyst^.^

142

Chapter 5

I
U

E
1
Q)

E 3

Lapse time / min


Figure 5.10

Cumulative volume of liquids produced in the thermal and catalytic degradation of HDPE at 430 "C in a batch reactor.20 Catalysts: ZSM-5, KFS-16 and amorphous silica-alumina (SA-1, SA-2). (Reprinted from J . Anal. Appl. Pyrol., 43, Y. Sakata, M.A. Uddin, A. Muto, Y. Kanada, K. Koizumi and K. Murata, page 15. 0 1997, with

permission from Elsevier Science) reactor at 430 "C over samples of amorphous silica-alumina, ZSM-5 zeolite and KFS-16. It can be seen that the rate of PE conversion over KFS-16 is similar to that of the amorphous silica-alumina and clearly superior to that of ZSM-5. The authors explain the activity of non-acidic KFS-16 by the stabilization of the radicals produced in the PE thermal degradation within the catalyst pores, which accelerates the degradation reactions. The absence of acid sites in KFS- 16 also caused a slower deactivation of this material compared with the acid catalysts. Figure 5.1 1 shows the liquid products obtained in successive cracking experiments over the same KFS-16 sample. Slight changes are observed from one experiment to another, the activity being in all cases superior to that of thermal degradation. Regeneration of the catalyst by calcination in air at 600 "C allowed most of the initial activity to be recovered. A completely different catalytic system for PE conversion in a fixed bed reactor has been studied by Uemichi et a1.2' The catalyst consisted of activated carbon impregnated with different transition metals (Pt, Fe, Mo, Zn, Co, Ni and Cu), which led to a bifunctional material with both cracking and dehydrogenation/hydrogenation activity. In all cases, high conversions were obtained and the main products were linear alkanes and aromatics, with little formation of

Catalytic Cracking and Reforming

143

12
10

8
6
4

J v

v
0

1 2 3 4

400

Therma
1st run 2nd r u n 3rd run

4th run

Regeneratec
500

100

200

300

600

Lapse time / min


Figure 5.11 Cumulative volume of liquids produced in the thermal and catalytic degradation of HDPE at 430 "C in a batch reactor.20 Catalysts: KFS-16 reused up to four times and then regenerated by calcination at 600 "C. (Reprinted from J. Anal. Appl. Pyrol., 43, Y. Sakata, M.A. Uddin, A. Muto, Y. Kanada, K. Koizumi and K. Murata, page 15. 0 1997, with permission from Elsevier Science)

isoalkanes and alkenes. The major effect of the metal incorporation on the activated carbon was to increase the selectivity towards aromatics and to decrease the formation of n-alkanes. The addition of the metals also caused a certain reduction in the overall conversion due to a higher coke deposition. The best yields in aromatics were obtained over Pt, Fe and Mo supported on activated carbon. The authors also investigated the effect of the space time (the ratio of feed mass flow to catalyst weight), a maximum in the aromatic yield being observed for all the catalysts (Figure 5.12). At low space times, the aromatic selectivity tends to zero, which shows that aromatic hydrocarbons are formed from decomposed fragments but not directly from PE itself. At higher space times, aromatics are converted through dealkylation and coking reactions, which explains the observed decrease in the aromatic yield. The activated carbon in this system is not a simple metal support, as can be concluded from the high activity obtained with this material alone. Moreover, impregnation of Pt and Fe over alumina and amorphous silica-alumina led to catalysts with lower yields of alkanes and aromatics, as they promote the formation of alkenes and isoalkenes. It is proposed that the initial hydrogen abstraction from the

144
00

Chapter 5

H U
v

40

u)
.r(

E
m
u4 0
4 P

rg

20

-W A

0
0

20

40

hO

W/F ( g c a t rnin/g

PE)

Figure 5.12 Relationship between contact time ( W / F )and aromatic yield in the catalytic degradation of P E in a fixed bed reactor over metal-supported activated carbons:2' 0 C, 0 Mo/C, 3 Ni/C, 0Zn/C, A Fe/C, A Cu/C, Pt/C, co/c. (Reprinted from J. Anal. Appl. Pyrol., 14, Y. Uemichi, Y. Makino and T. Kanazuka, page 33 1. 0 1989, with permission from Elsevier Science)

polymer takes place on the activated carbon surface, and the resulting hydrogen atoms migrate to the metal sites where they are easily desorbed. In contrast with the great research effort put into the preparation, characterization and testing of different catalytic systems for the conversion of plastics, only a few studies have been published aimed at the design and operation of feasible catalytic reactors. Thus, most of the studies previously described use simple batch or fixed bed reactors, in spite of the problems associated with the high viscosity and low heat transfer of the molten plastics. Some authors have also investigated the catalytic degradation of PE in fluidized bed reactors.22923 Scott et a1.22 used an activated carbon bed for the degradation of LLDPE, obtaining large amounts of gaseous products, probably as a result of the high temperatures of operation (500-785 "C). Likewise, Sharratt et al.23 have recently developed a fluidized bed system for the catalytic conversion of plastics. The catalyst is pelletized to give particles with sizes in the range 75-180 pm and is fluidized by passing through a nitrogen stream. Once the reaction temperature is reached, the polymer is added as particles with diameters between 75 and 250 pm. After melting, the polymer wets the catalyst surface, being pulled into the catalyst macropores by capillary action. The reactor works with low polymer/catalyst ratios between 0.2 and 1.0. This system has been used for the catalytic degradation of HDPE over ZSM-5 zeolite in the temperature range 290430C. The main products

Catalytic Cracking and Reforming

145

obtained were gaseous hydrocarbons with yields up to 70 wt%, which increased with the reaction temperature. At 360"C, a conversion of over 90% was obtained after just 15 min of reaction. The gases were rich in C3-C4 olefins, whereas C5 olefins were the predominant hydrocarbons in the liquid fraction. The aromatic content of the latter depended strongly on the reaction temperature, with a maximum at 390C. These results show the possibility of obtaining high yields of C3-C5 olefins, with a variety of applications in chemical synthesis, by catalytic cracking at temperatures about 350 "C. As was described in Chapter 4, high conversion of PE into olefinic gases is only achieved by thermal degradation when working at high temperatures, usually above 700 "C. Another alternative for the feedstock recycling of plastic wastes by catalytic degradation consists of the direct feeding of plastics dissolved in gas oil fractions to FCC (fluidized catalytic cracking) refinery units. Using this method, the conversion of PE blended with a vacuum gas oil (VGO) has been investigated in a fixed bed reactor at 510C over a commercial FCC catalyst.24 It was concluded that the amount of PE in the blend is critical. The cracking of a blend containing 5 wt% of HDPE produced little or no gasoline, but only gas and coke. However, as the HDPE concentration was increased to 10 wt%, a significant amount of gasoline was produced. Further increases in the PE content of the raw mixture yield blends which are too viscous for feeding to the FCC unit.

Polypropylene
The catalytic conversion of PP has been investigated using catalysts, reactors and conditions very similar to those described above for PE. A decrease in the activation energy of PP degradation in the presence of catalysts has also been observed by Audisio et al.25 based on TGA measurements, although this effect was not as remarkable as in the case of PE. Compared with the activation energy corresponding to the thermal cracking of PP (122.7 kJ mol-I), the catalytic conversion over HY and REY led to values of 108.9 and 117.7 kJ mol-'. A greater reduction was observed over an amorphous silicaalumina catalyst, with an activation energy of 99.2 kJ mol-'. The authors propose that this relatively small effect of the catalyst in PP degradation is probably associated with the presence of a methyl substituent in the polymer chain, and therefore of a high concentration of tertiary carbons, which makes thermal cracking easier, weakening the catalyst effect. The mechanism of PP degradation over amorphous silica-alumina catalysts has been investigated by Ishihara et al.13 in a batch reactor. The authors propose a sequential pathway, as in the case of PE degradation. Under the same reaction conditions, PP is more rapidly converted than PE. Thus at 280 "C, 1 h of reaction, and catalyst/polymer ratio = 1, a PP conversion of 43% is obtained while just 20% of PE is degraded into volatile products. The higher reactivity of PP is due to its higher branching frequency, which favours cracking reactions. . also ~ ~studied the degradation of PP over amorphous Sakata et ~ 1 have

146

Chapter 5

silica-alumina catalysts. At 380 "C the products obtained consisted of 68.8% liquid hydrocarbons, 24.8% gases and 6.4% residue. The gases formed were mainly butene and propylene, while the liquids contained a complex mixture of hydrocarbons in the CS-Cl6 range. Compared with the results obtained in the thermal cracking and in two-step (thermal-catalytic) treatments, the higher PP degradation rate was obtained by direct contact between the polymer and the catalyst. This result confirms that the solid acid catalyst in contact with the melted plastic promotes the polymer degradation, participating in the earlier stages of the cracking mechanism. A detailed characterization of the products formed in the catalytic degradation of PP over amorphous silica-alumina and CaX zeolite has been performed by Uemichi et ~ l . , * using ~ a fixed bed reactor at temperatures between 470 and 526C. For both catalysts the gases produced were mainly isobutene and isobutane, a higher proportion of the alkane being obtained over the zeolite catalyst. Similarly, the liquid fraction consisted mainly of isoalkanes and alkenes, and minor amounts of linear alkanes and aromatics. Most of the branching was due to the presence of side methyl groups. The products formed over CaX were more saturated, with lower amounts of alkenes and aromatics. Compared with thermal degradation, the products derived from the catalytic conversion contained greater quantities of aromatics and monomethylbranched hydrocarbons, especially those with branching at the odd number. CaX was deactivated faster than the amorphous silica-alumina, due to the increased coke deposition that takes place over the zeolite. Zhao et a1.2ghave studied PP catalytic degradation over different zeolites by means of TGA measurements. The following order of activity was observed: zeolite Y > mordenite > zeolite L. PP conversion over HY led to hydrocarbons concentrated in the range C&9, and some new compounds with cyclic structures were found compared to the thermal degradation. The limitations of microporous catalysts for the conversion of PP have recently been pointed out by Aguado et al.,7 by comparing the activity and product distribution obtained over ZSM-5 zeolite, MCM-41 and amorphous silica-alumina. Figure 5.13 shows the conversion and the product distribution obtained in a batch reactor. It is remarkable that the ZSM-5 zeolite led to a conversion just slightly superior to that of the thermal degradation, in contrast with the high activity obtained over this catalyst in the conversion of both HDPE and LDPE. On the contrary, a significant activity was observed over the silica-alumina catalyst, while MCM-41 led to a total conversion of the polymer into mainly gasoline and middle distillate fractions. The low activity exhibited by the ZSM-5 sample in the PP conversion has been related to the steric hindrance for the access of PP oligomers into the zeolite channel system, due to the presence of side methyl groups on half of the backbone carbons. This conclusion was supported by the results of the molecular simulation of the adsorption of PP oligomers into the ZSM-5 structure, which showed that they cannot be accommodated within the pores and cavities of this zeolite. Other possibilities for the catalytic conversion of PP over zeolites are to use zeolitic structures with a larger pore size or to increase the proportion of

Catalytic Cracking and Reforming

147

Figure 5.13 Conversion and product distribution obtained in PP catalytic cracking at 400C in a batch reactor over ZSM-5, MCM-41, and amorphous silicaalumina.'

Figure 5.14 Conversion obtained in the thermal and P P catalytic cracking at 400 "C in a batch reactor.29

external surface area through a reduction in the zeolite crystal size. Figure 5.14 compares the activity obtained during the degradation of PP over different zeolites.29 Increasing the pore size (zeolites USY and Beta) causes a certain improvement in the catalytic activity, although it is still very low compared to that of mesoporous MCM-41, which indicates that strong steric and/or diffusional constraints hinder the PP conversion. However, a significant increase in the conversion is obtained when using ZSM-5 with very small crystals as ~atalyst.~' In this case, the initial PP cracking takes place over the

148

Chapter 5

external surface area to yield small fragments that can enter the zeolite pores, where they are converted through further cracking, isomerization and aromatization reactions. In addition to amorphous silica-alumina and zeolites, activated carbons have also been used as catalysts for PP cracking. Thus, Nakamura and Fujimoto3' have studied the degradation of PP over a catalyst of Fe supported on activated carbon in a batch reactor at temperatures between 380 and 400C. It is proposed that the cracking reaction is initiated by free radicals present on the activated carbon surface through the abstraction of hydrogen from the tertiary carbons of the PP chains. The addition of Fe on activated carbon caused a slight increase in the activity compared to the support alone. However, incorporation of small amounts of CS2 or H2S led to a significant increase in the catalytic activity of the Fe/activated carbon system and to a reduction in the production of gases. The authors propose that hydrogen from H2S is abstracted by the radicals generated by thermal degradation of PP to form a stable hydrocarbon and an HS' radical, which avoids the consecutive cracking of the hydrocarbon radical and the formation of overcracked products (gases). As a consequence, the main products obtained by PP degradation with this system are naphtha, kerosene and gas oil fractions.

Polystyrene
In contrast with PE and PP, thermal degradation of PS takes place at relatively low temperatures with a high yield of the raw monomer. Accordingly, in most studies on the catalytic decomposition of PS, the polymer is simultaneously degraded by both radical and ionic pathways. . ~ investigated ~ the degradation of PS over different Audisio et ~ 1 have catalysts: alumina, silica, amorphous silica-alumina, and zeolites HY and REY. From TGA measurements, the activation energy of the thermal PS cracking (277 kJ mol-') was compared with that of the catalytic process. In this case, amorphous silica-alumina samples caused the greatest reduction in the activation energy, to 45.6 kJ mol-', followed by the zeolites with values of 96.1 and 115.8 kJ mol-' over HY and REY, respectively. Moreover, amorphous silica-alumina catalysts allowed the degradation temperature to be reduced to about 270C. While the major product of the thermal degradation of PS is styrene, only minor amounts of the monomer were obtained in the catalytic experiments, the main products being benzene, ethylbenzene, a-methylstyrene, toluene, isopropylbenzene and indane compounds. The authors propose that the catalytic degradation is initiated by proton addition to PS, which generates ionic species with the positive charge on the benzylic substituent of the polymer chain. Subsequent p-scission, isomerization and cyclization reactions lead to the formation of the different aromatic products. However, other authors have proposed that the primary product of the PS catalytic cracking is styrene, as in thermal cracking, which is further converted into ethylbenzene, toluene, benzene, etc. on the acid sites of the catalyst. De la

Cataly tic

149

Conversion (wt.-%)

Figure 5.15

Selectivities of benzene (O), toluene ( O ) ,ethylbenzene (A), and cracking products (V) as a function o f conversion at 550 "C in the catalytic treatment o f PS (dashed line and open symbols) and styrene (filled line and closed symbols) over a commercial FCC (Reprinted with permission from G. de la Puente, J.M. Arandes and U.A. Sedran, Ind. Eng. Chem. Res., 36,4530. 0 1997 ACS)

Puente et have studied the conversion of PS dissolved in benzene in a fluidized bed reactor over a commercial FCC catalyst. The product distribution obtained in the catalytic degradation of PS was compared to that obtained in styrene conversion (Figure 5.15). The same relationship between the conversion and the selectivity towards the different products was observed in both PS and styrene catalytic conversion at 550C, suggesting that styrene is also the primary product in the catalytic PS cracking. The authors proposed a mechanism to explain the formation of the main products from styrene. A rapid deactivation of acid catalysts has been observed by Uemichi et al." during PS cracking over HY, H-mordenite, ZSM-5 and silica-alumina, which is related to the high concentration of aromatic hydrocarbons. Likewise, Serrano et al.34 have observed that, depending on the reaction conditions, catalytic degradation of PS may lead to lower conversions than a simple thermal treatment. This is due to the existence of crosslinking reactions between the polymer chains, which are promoted by the acid sites of the catalysts and yield a residue of low reactivity. The extent of crosslinking versus cracking increases with the acid strength of the catalyst. Thus, very low PS conversions were obtained in a batch reactor over ZSM-5 zeolite. Taking into account the pore size of this zeolite and the cross-sectional diameter of the PS molecules with side benzylic groups, it is concluded that crosslinking takes place over the external zeolite surface.

150

Chapter 5

4 Catalytic Conversion of Plastic Mixtures and Rubber Wastes


This section describes the different processes that have been patented for the catalytic conversion of plastic mixtures without any previous thermal treatment. In many cases, the authors claim that the process is also successful in the degradation of rubber wastes or plastic and rubber mixtures. While many studies have been carried out aimed at the feedstock recycling of rubber wastes by pyrolysis and hydrogenation processes (see Chapters 5 and 7), little information is found on the catalytic cracking and reforming of rubber alone. L a r ~ e n has ~~ disclosed that waste rubber, such as used tyres, can be degraded in the presence of molten salt catalysts with properties as Lewis acids, such as zinc chloride, tin chloride and antimony iodide. The decomposition proceeds at temperatures between 380 and 500 "C to yield gases, oil and a residue, in proportions similar to those obtained by simple thermal decomposition. Wingfield et aZ.36have developed a process for the catalytic treatment of rubber and plastic waste by reaction in the presence of Zn and Cu salts (chlorides or carbonates). Examples are provided for the conversion of automobile shredder waste, containing a complex mixture of polymers (rubber, polyurethane, polyester, PP, PVC, ABS, etc.), at 550C to produce char and oil fractions, the latter containing about 0.5 wt% of S and N. In another patent,37these authors disclose that plastic and rubber wastes can also be degraded in the presence of basic salt catalysts, such as sodium carbonate. Likewise, Platz3*has reported the catalytic conversion of plastics and rubber by contact with a molten MgC12/A1C13catalyst, whereas Butcher39has reported on the degradation of polymeric materials over molten mixtures of a basic salt (NaOH or KOH) and a Cu source, mainly metallic Cu and CuO. Processes involving the use of solid acid catalysts have also been patented. According to Chen and Yan?' plastic and/or rubber wastes are first subjected to a size reduction step, followed by separation of any metals present and washing to remove any non-plastic material such as paper, labels, etc. Subsequently, the polymer wastes are dissolved or dispersed in a petroleum oil, with a high content of polycyclic aromatic compounds at 300 "C, and catalytically transformed in an FCC reactor at temperatures of about 500C. Details are given for the conversion of different wastes: used whole tyres, PE bags and PS foam. Saito and Nanba4' have developed several processes for the catalytic conversion of plastic wastes, one of them being shown in Figure 5.16. The plastic scraps are fed together with the catalyst to a first reaction vessel by means of a screw placed above it. The catalyst/plastic ratio is typically in the range 0.050.1. In this reactor the plastics are melted and subjected to catalytic cracking under agitation at temperatures between 400 and 470 "C. The volatile products leaving the first vessel can be separated into fractions or alternatively they may pass through a second catalytic reactor (fixed bed) to further modify the product distribution. The use of zeolitic catalysts is recommended. In the first reactor, the spent catalyst can be continuously removed through a screw feeder located

Catalytic Cracking and Reforming

151

Figure 5.16

Schematic diagram of an apparatus for the thermal andlor catalytic degradation of plastic scrap^.^' (0 K. Saito and M. Nanba, US Patent 4 584421,1986)

in the bottom of the vessel and regenerated by treatment at 500 "C in a separate unit . The catalytic degradation of a number of polymer mixtures commonly found in plastic wastes has been reported by Evans and using two-step processes at different temperatures in order to isolate the decomposition of each polymer and to avoid mixing of the corresponding degradation products. Thus, a mixture of nylon-6 and PP is first degraded at 293C over a KOH/alumina catalyst, which leads to the recovery of caprolactam from the decomposition of nylon-6, and then it is subjected to a temperature of 397 "C to recover the products derived from the PP degradation. Similar treatments are described for the catalytic conversion of other polymer mixtures: PET + HDPE, PET + PS + PE, PVC + PU, polyphenylene oxide (PPO) + PS, PC + ABS, etc. It is proposed that when PVC is present in the starting mixture, the HCl evolved may catalyse the decomposition of the other polymers.

Conversion of Plastics by a Combination of Thermal and Catalytic Treatments

Two-step processes based on the combination of a previous thermal treatment and a subsequent catalytic conversion, are widely used to facilitate the plastic flow and its mixing and contact with the catalyst. These two treatments can be carried out in two zones of the same reactor, in two different reactors of the same plant, or even in two different plants. Ohkita et ai.43have used a two-zone reactor for the catalytic degradation of PE at 400 "C over amorphous silica-alumina and ZSM-5 zeolite. The polyolefin

152

Chapter 5

is placed in the bottom of a stainless steel reactor, whereas the catalyst is placed in the middle part of the tube, so that the vapour generated by the thermal decomposition of PE passes through the catalyst bed. Conversion over ZSM-5 led to a higher proportion of gases, mainly C3 and C4, compared to amorphous silica-alumina, whereas the oils were rich in aromatic hydrocarbons. Figure 5.17 illustrates a GC analysis of the oils produced over this zeolite, with assignments of the peaks corresponding to aromatic compounds. Over amorphous silica-alumina, the concentration of aromatics in the oils were observed to increase with the concentration of acid sites, which can be controlled by varying the aluminium content of the catalyst. A two-zone reactor has also been used for the conversion of PS at 350 "Cover both acid and basic catalysts.44 Considerable amounts of benzene and ethylbenzene were formed over acid catalysts such as ZSM-5 and silica-alumina, showing that the styrene produced in the thermal step is further converted in the catalyst bed by cracking and hydrogenation. However, with basic catalysts the fraction of styrene in the oils increased, values up to 75 wt% being observed. Figure 5.18 illustrates the yields of styrene monomer and dimer produced over the different catalysts investigated. The degradation was faster over basic oxides, mainly with BaO and K20, compared to the acid catalysts, which is assigned to the higher conversion rate of vaporized PS fragments and oligomers on the basic sites. It is proposed that PS degradation on solid bases is initiated through the formation of carboanions by proton abstraction from PS on the basic sites.

Figure 5.17 GC-MS analysis of the oils produced by thermal-catalytic PE degradation at 400 "C in a two-zone reactor with assignments of the peaks corresponding to aromatic hydrocarbon^.^^ (Reprinted with permission from H. Ohkita, R. Nishiyama, Y. Tochihara, T. Mizushima, N. Kakuta, Y. Morioka, A. Ueno, Y. Namiki, S. Tanifuji, H. Katoh, H. Sunazuka, R. Nakayama and T. Kuroyanagi, Ind. Eng. Chem. Res., 32,3112. 0 1993 ACS)

Catalytic Cracking and Reforming

153

P .P

>

I
catalyst

HZSMS

Figure 5.18

Yields of styrene monomer and dimer obtained in the thermal-catalytic f P S at 350C in a two-zone reactor over various metal oxide degradation o catalysts. (Reprinted with permission from Z. Zhang, T. Hirose, S. Nishio, Y. Morioka, N. Azuma, A. Ueno, H. Ohkita and M. Okada, Ind. Eng. Chem. Res., 34,4514. 0 1995 ACS)

Songip et ~ 1have . studied ~ ~ the catalytic reforming of a heavy oil obtained by thermal degradation of PE at 450 "C. Prior to the catalytic treatment, the oil was distilled at 300C to remove the residue (20-30 wt%) and reduce its viscosity. Analysis of the oil showed it contained mostly paraffinic hydrocarbons. Figure 5.19 shows the product distribution obtained in the catalytic conversion of this oil in a fixed bed reactor at 400C over different acid catalysts. ZSM-5 zeolite with Si/Al = 65 showed the highest activity, followed by zeolites REY and HY. Amorphous silica-alumina and a sample of ZSM-5 with very low A1 content (%/A1 = 100) were the least active catalysts. Large differences were observed between the product distributions corresponding to the different zeolites. ZSM5 with Si/Al = 65 gave the largest yield of gaseous products (69 wt%) and the lowest amount of gasoline, this fraction being even lower than in the raw oil. The gases were rich in C3 and C4 olefins. On the contrary, a high proportion of gasoline fraction was produced over REY zeolite. Coke deposition was more severe on HY and REY zeolites than over ZSM-5, which caused a faster deactivation of the former catalysts. The RON values corresponding to raw oil and the liquid products obtained after the catalytic treatments are shown in Figure 5.20. It is remarkable that the RON of the starting oil, produced by PE thermal degradation, is almost zero due to its high content of linear paraffins. As far as the gasoline fractions produced by catalytic reforming are concerned, the highest RON values correspond to the products formed over REY and ZSM-5 zeolites, followed by the gasoline obtained over HY.

154

GASOLINE

,~
G A . S ,
20%
40%

Chapter 5
COT. H;?Wp

HY

REY
SILICA ALUMINA
ZSM-5 (1000)
ZSMd (65) PE OIL

0%

60%

80%

100%

YIELD, W

Y o

Figure 5.19 Product distribution obtained in the catalytic conversion of a PE pyrolysis oil in aJixed bed reactor (T = 400 C, WHSV = 1, t = 3 h).45

SILICA ALUMINA

ZSM-5 (1000)
ZSM-5 (65)

PEOIL

]
I
I

20

40

60

80

RON VALUE

Figure 5.20 RON values of the gasoline fraction obtained in the catalytic conversion of a PEpyrolysis oil in afixed bed reactor (T = 400 C, WHSV = 1, t = 3 h).45

In further works, Songip et al.46*47 have studied the effect of the reaction conditions and the kinetics of the catalytic reforming of the PE oil over REY zeolites in a fixed bed reactor. Figures 5,21(A)-(C) show the yields of gasoline, gases and coke versus the oil conversion, obtained in experiments carried out at different temperatures and space times. All the data corresponding to the gas yield lie on the same curve, indicating that the temperature has no effect on the gas production, but that it depends mainly on the oil conversion. When more than 80% of the starting oil has been converted, the gas yield grows exponen-

Catalytic Cracking and Reforming


60
50
c

155

rn
x

40

.-

30 20
10
A'

0
100

20

40 60 80 Conversion of hc3vy o i l [%I


I
I

100

z
0 .2

80

+ 573K - O 623K
A ZUK

60
40

20 0

0
1.2

20

40

60
I

80
I

100

Convenion of heavy oil [wt%)

9-"
I

+ 573 K
0 623K
673 K
A 723K

0 . 9
0.6

. .

.A

623K

C)

20 40 60 80 Conversion of hcsvy oil [%I

100

Figure 5.21 Product distribution obtained in the catalytic conversion of a PE pyrolysis oil over REY zeolite:46(A) gasoline yield, (B) gas yield, ( C ) coke yield. (Reprinted with permission from A.R. Songip, T. Masuda, H. Kuwahara and K. Hashimoto, Energy Fuels, 8, 136. 0 1994 ACS)

tially. At constant temperature, the gasoline yield exhibits a maximum oil conversion of about 8O%, which suggests that the gasoline fraction is initially formed by cracking of the heavy oil and undergoes subsequent cracking to yield gaseous hydrocarbons. Moreover, the gasoline yield also exhibits a maximum with respect to temperature at about 400C. Coke deposition on the catalyst increases with the oil conversion, but is reduced by increasing the reaction temperature. The variation in gasoline quality with the temperature is shown in Figure 5.22.46Below 400 "C,the RON value increases with the temperature due to the increase in the production of isoparaffins and the reduction in the content of linear paraffins. However, at higher temperatures, cracking of the isoparaffins takes place, leading to a certain reduction in the RON value.

156 120
Y

Chapter 5

RON

550

600

650 700 Temperature [K]

750

Figure 5.22 RON value and composition of the gasoline fraction produced in the catalytic conversion of a P E pyrolysis oil over R E Y zeolite, as functions of the temperat~re.~~ (Reprinted with permission from A.R. Songip, T. Masuda, H. Kuwahara and K. Hashimoto, Energy Fuels, 8, 136. 0 1994 ACS)

A number of two-step processes (thermal and catalytic treatments) have been reported in the patent literature for the conversion of plastic wastes. Fukuda et al.48have proposed the thermal degradation of polyolefinic plastics in a stirred reactor at 420-470 "C, the volatile products being subsequently passed through a fixed bed reactor at 250-340C, containing a ZSM-5 zeolite catalyst. The process developed by Lechert et af.49also consists of a two-reactor system: plastics are first pyrolysed above 600 "C in a sand fluidized bed reactor, and the gases produced are catalytically converted over ZSM-5 zeolite at 350-410 "C in a fixed bed reactor to increase the overall yield of liquid products. An interesting process is that developed jointly by Fuji Recycle and Mobil Oil for the treatment of plastic wastes containing PE,PP and PS (Figure 5.23).5015' The waste plastics are crushed and washed to remove impurities (dirt, paper, etc.). Flotation in water is used to segregate the different polymers present in the waste and to separate PVC and PET from PE,PP and PS.The resulting plastic mixture is warmed at 250 "C and introduced into the melting vessel at 300 "C by means of a heated extruder. The mixture is then transferred into the thermal cracking vessel where it is decomposed at about 400C. The gases generated pass through the catalytic reactor, containing ZSM-5 zeolite, to be transformed into higher value hydrocarbons. Final liquid and gas fractions are separated by condensation, the gases being used as in-house fuel. A part of the product of the thermal cracking reactor is returned to the melting vessel, a settler being available in the connecting pipeline to remove coke and other impurities. Processing PE and PP in this system typically yields 80% liquids, 15% gases and 5% residue. The liquid is basically composed of 50% gasoline, 25% kerosene and 25% gas oil. When converting PS, the products are mainly aromatic hydrocarbons, with a high content of ethylbenzene, toluene and benzene.

Catalytic Cracking and Reforming

K.O. POT HOPPER

CATALYTIC REACTOR

157

RECYCLE FURNACE

CRACKED OIL TO PRODUCT TANK

HEAT TRANSFER OIL FURNACE

SETTLER CRACKED GAS

Figure 5.23 Flow diagram of the Fuji Recycle-Mobil Oil rocess for the conversion of plastic wastes by thermal-catalytic treatments.

6 Summary
Catalytic cracking of plastic wastes offers a number of advantages compared with simple thermal degradation. The presence of catalysts greatly increases the polymer cracking rate, which allows the use of lower temperatures and/or shorter reaction times. Moreover, by a suitable choice of catalysts it is possible to direct the plastic degradation to the formation of different products: olefinic gases, gasoline fractions and middle distillates. Thus, the use of catalysts with strong acidity in fluidized bed reactors has allowed high yields of gases rich in C3 and C4 olefins to be obtained, yields of up to 70% being reported at temperatures of about 400 "C. Likewise, the liquid hydrocarbons obtained in the gasoline range are high quality products because they contain mainly branched alkanes and aromatics, in contrast with the thermal degradation oils, which are formed by both linear alkanes and a-olefins. However, these methods of plastic recycling are mainly limited to the degradation of polyolefinic plastics, because the presence in the feed of C1 or N-containing polymers may lead to a poisoning of the catalyst active sites. Likewise, the inorganic fillers and contaminants contained in the raw wastes tend to remain with the solid catalysts, which means that further separation steps are necessary. The catalysts commonly used to promote plastic degradation are a variety of acid solids such as amorphous silica-alumina, different types of zeolites, mesoporous aluminosilicates (MCM-41), sulfated zirconia, etc. Interesting results have also been obtained in polymer cracking over activated carbons

158

Chapter 5

(alone or impregnated with transition metals), silica-based mesoporous materials (KFS- 16), and basic solids. The activity and product distribution obtained in the PE, PP and PS cracking over these materials depends mostly on the main catalyst properties: type, strength and concentration of active sites, available surface area, average pore diameter, pore size distribution, etc. The best catalysts for the conversion of PE and PP are solid acids, although they usually undergo a significant deactivation due to coke deposition. However, in most cases the initial activity can be restored by burning off the coke deposits, On the other hand, PS is degraded faster on basic catalysts, because acid materials promote crosslinking reactions between the polymer chains. The intense research effort carried out into the study of catalyst properties for the conversion of plastic wastes is in contrast with the few studies that have addressed reactor design. Thus, most of the studies use batch or simple fixed bed reactors despite the heat transfer and flow problems associated with the low thermal conductivity and high viscosity of the molten plastics. Various alternatives have been proposed to solve these problems: the use of fluidized bed reactors, dissolution of the plastics in heavy oil fractions previously fed into the reactor, and a combination of thermal and catalytic treatments. However, all these processes present a number of difficulties, which makes further work on the reactor design necessary. Likewise, it is convenient to check the behaviour of the catalysts when working with real plastic wastes rather than simply with model polymers, in order to ascertain the effect of the various impurities and contaminants the wastes may contain on the catalyst activity and stability.

7 References
1 H. Robson, Microporous Mater., 1998,22(46), 495. 2 R. Szostak, Molecular Sieves. Principles of Synthesis and Identification, Van Nostrand Reinhold, New York, 1989. 3 W.O. Haag, Stud. Su$. Sci. Catal., 1994,84, 1375. 4 C.T. Kresge, M.E. Leonowicz, W.J. Roth, J.C. Vartuli, and J.S. Beck, Nature, 1992, 359, 710. 5 J.S. Beck, J.C. Vartuli, W.J. Roth, M.E. Leonowicz, C.T. Kresge, K.D. Schmitt, C.T.-W. Chu, D.H. Olson, E.W. Sheppard, S.B. McCullen, J.B. Higgins, and J.L. Schlenker, J . Am. Chem. SOC., 1992, 114, 10834. 6 J. Aguado, D.P. Serrano, M.D. Romero, and J.M. Escola, Chem. Commun., 1996, 1,725. 7 J. Aguado, J.L. Sotelo, D.P. Serrano, J.A. Calles, and J.M. Escola, Energy Fuels, 1997,11, 1225. 8 J. Kodaira, Z. Osawa, and H. Ando, J . Chem. SOC. Jpn, Chem. Ind. Chem., 1977,12, 1892. 9 S.R. Ivanova, E.F. Gumerova, K.S. Minsker, G.E. Zaikov, and A.A. Berlin, Prog. Polym. Sci., 1990, 15, 193. 10 P.L. Beltrame, P. Carniti, G. Audisio, and F. Bertini, Polym. Degrad. Stab., 1989,26, 209. I 1 R.C. Mordi, R. Fields, and J. Dwyer, J . Anal. Appl. Pyrof., 1994,29,45. 12 Y. Ishihara, H. Nambu, K. Saido, T. Ikemura, and T. Takesue, Polymer, 1992, 33(16), 3482.

Catalytic Cracking and Reforming

159

13 Y. Ishihara, H. Nambu, T. Ikemura, and T. Takesue, J. Appl. Polym. Sci., 1989, 38, 1491. 14 Y. Ishihara, H. Nambu, T. Ikemura, and T. Takesue, Fuel, 1990,69,978. 15 M.A. Uddin, K. Koizumi, K. Murata, Y. Sakata, Polym. Degrad. Stab., 1997,56,37. Jpn, 16 Y. Ishihara, H. Nambu, K. Saido, T. Ikemura, and T. Takesue, Bull. Chem. SOC. 1991,64,3585. 17 Y.-H. Lin, P.N. Sharratt, A.A. Garforth, and J. Dwyer, Thermochim. Acta, 1997, 294,45. 18 Y. Uemichi, M. Hattori, T. Itoh, J. Nakamura, and M. Sugioka, Ind. Eng. Chem. Res., 1998,37, 867. 19 R. Lin and R.L. White, J. Appl. Polym. Sci., 1995,544, 1151. 20 Y. Sakata, M.A. Uddin, A. Muto, Y. Kanada, K. Koizumi, and K. Murata, J. Anal. Appl. Pyrol., 1997,43, 15. 21 Y. Uemichi, Y. Makino, and T. Kanazuka, J. Anal. Appl. Pyrol., 1989,14, 331. 22 D.S. Scott, S.R. Czernik, J. Piskorz, and D.St.A.G. Radlein, Energy Fuels, 1990, 4, 407. 23 P.N. Sharratt, Y.-H. Lin, A.A. Garforth, and J. Dwyer, Ind. Eng. Chem. Res., 1997, 36, 51 18. 24 S.H. Ng, Energy Fuels, 1995,9,216. 25 G. Audisio, F. Bertini, P.L. Beltrame, and P. Carniti, Makromol. Chem., Macromol. Symp., 1992,57, 191. 26 Y. Sakata, M.A. Uddin, K. Koizumi, and K. Murata, Chem. Lett., 1996, 1, 245. 27 Y. Uemichi, Y. Kashiwaya, M. Tsukidate, A. Ayame, and H. Kanoh, Bull. Chem. SOC. Jpn, 1983,56,2768. 28 W. Zhao, S. Hasegawa, J. Fujita, F. Yoshii, T. Sasaki, K. Makkuuchi, J. Sun, and S . Nishimoto, Polym. Degrad. Stab., 1996,53, 129. 29 D.P. Serrano, Proceedings of the 5th FEZA Euroworkshop on Microporous Materials: their Application in Catalysis, Gandia, Spain, 1998. 30 J.L. Sotelo, R. van Grieken, J. Aguado, D.P. Serrano, J.M. Escola, and J.M. Menendez, Proceedings of the 12th International Zeolite Conference, Baltimore, 1998, ed. M.M.J. Treacey, B.K. Marcus, M.E. Bisher and J.B. Higgins, Materials Research Society, 1999, Vol. 11, p. 1441. 31 I. Nakamura and K. Fujimoto, Catal. Today, 1996,27, 175. 32 G. Audisio, F. Bertini, P.L. Beltrame, and P. Carniti, Polym. Degrad. Stab., 1990,29, 191. 33 G. de la Puente, J.M. Arandes, and U.A. Sedran, Ind. Eng. Chem. Res., 1997, 36, 4530. 34 D.P. Serrano, J. Aguado, and J.M. Escola (submitted to Appl. Catal. B: Environ.). 35 J.W. Larsen, US Patent 3 996022, 1976. 36 R.C. Wingfield, Jr., J. Braslaw, and R.L. Gealer, US Patent 4 458 095, 1984. 37 R.C. Wingfield, Jr., J. Braslaw, and R.L. Gealer, US Patent 4 515 659, 1985. 38 G.M. Platz, US Patent 5 504 267, 1996. 39 J.A. Butcher, Jr., US Patent 5 315 055, 1994. 40 N.Y. Chen and T.-Y. Yan, US Patent 4 108 730, 1978. 41 K. Saito and M. Nanba, US Patent 4 584 421, 1986. 42 R.J. Evans and H.L. Chum, US Patent 5 321 174, 1994. 43 H. Ohkita, R. Nishiyama, Y. Tochihara, T. Mizushima, N. Kakuta, Y. Morioka, A. Ueno, Y. Namiki, S. Tanifuji, H. Katoh, H. Sunazuka, R. Nakayama, and T. Kuroyanagi, Ind. Eng. Chem. Res., 1993,32, 3 1 12. 44 Z. Zhang, T. Hirose, S. Nishio, Y. Morioka, N. Azuma, A. Ueno, H. Ohkita, and M. Okada, Ind. Eng. Chem. Res., 1995,34,4514. 45 A.R. Songip, T. Masuda, H. Kuwahara, and K. Hashimoto, Appl. Catal. B: Environ., 1993,2, 153. 46 A.R. Songip, T. Masuda, H. Kuwahara, and K. Hashimoto, Energy Fuels, 1994, 8, 136.

160

Chapter 5

47 A.R. Songip, T. Masuda, H. Kuwahara, and H. Hashimoto, Energy Fuels, 1994, 8, 131. 48 T. Fukuda, K. Saito, S. Suzuki, H. Sato, and T. Hirota, US Patent 4851 601, 1989. 49 H. Lechert, V. Woebs-Gosch, S. Qun, W. Kaminsky, and H. Sinn, US Patent 4 871 426, 1989. 50 T. Tachibana, Eur. Patent 675 189A1, 1995. 51 T. Hirota and F.N. Fagan, Makromol. Chem., Macromol. Symp., 1992,57, 161.

CHAPTER 6

Hydrogenation
1 Introduction
Hydrogenation of plastic and rubber wastes is a potentially interesting alternative for breaking down the polymer chains. Compared to treatments in the absence of hydrogen, hydrogenation leads to the formation of highly saturated products, avoiding the presence of olefins in the liquid fractions, which favours their use as fuels without further treatment. Moreover, hydrogen promotes the removal of heteroatoms (Cl, N and S) that may be present in the polymeric wastes. However, hydrogenation suffers from several drawbacks, mainly the cost of hydrogen and the need to operate under high pressures. Although some non-catalytic hydrogenation processes have been reported, most of them require the presence of a catalyst to promote hydrogen addition reactions. Bifunctional catalysts are preferred, incorporating both cracking and hydrogenation/dehydrogenation activities. A typical catalyst includes transition metals (Pt, Ni, Mo, Fe, etc.) supported on acid solids, such as alumina, amorphous silica-alumina, zeolites, sulfated zirconia, etc. These materials are similar to those used in the catalytic cracking and reforming of plastic wastes described in Chapter 5, although in this case the incorporation of metals, metal oxides or metal sulfides provides them with hydrogenation/dehydrogenation activity. In recent years, a lot of research effort has been focused on the coprocessing of coal and rubber or plastic wastes under hydrogen atmospheres, based on the possible role of the polymers as hydrogen donor for coal. Therefore, in this chapter the description of plastic and rubber hydrogenation processes is followed by sections dealing with their coprocessing with coal.

2 Hydrocracking of Plastics
Plastic hydrogenation has recently been used for the conversion of various polymers: PE, PP, PS, PET and mixtures. The reactions are usually conducted in autoclaves at temperatures around 400 "C, under pressures of cold hydrogen up to 150 atm, and in some cases in the presence of solvents. Enhanced activities are obtained by means of a variety of catalysts.
161

162

Chapter 6

Taghiei et al.' have studied the liquefaction of PE, PP, PET and a mixed waste plastic (MWP) at temperatures between 420 and 450 "C in the presence of hydrogen (54 atm of cold hydrogen) and using tetralin and a waste oil as solvents. The results of the thermal degradation were compared with those using two different catalysts: ZSM-5 zeolite and a highly dispersed Fe catalyst (ferrihydrite treated with citric acid). For the conversion of PE, ZSM-5 zeolite was significantly more active than both the Fe-based catalyst and the thermal treatment, leading to oil yields in the range 80-98%. As can be seen in Figure 6.1, ZSM-5 was also highly effective in the conversion of other plastic wastes. The differences between the catalysts were appreciably reduced in the liquefaction of PP, since due to the higher reactivity of this polyolefin, oil yields over 80% are obtained just by simple thermal degradation at 420 "C. Likewise, no significant differences were observed between the thermal and catalytic degradation of PET. While the formation of gaseous products was negligible in the PE and PP conversions, around 13% of gases were produced from PET. In a further work,2 the authors studied the processing of plastic wastes in mixtures of tetralin and used automotive oil. Figure 6.2 shows the results obtained with different waste oil/tetralin ratios. ZSM-5 was again more active than the ferrihydrite-based catalyst, while the latter led to oil yields very close to those of the thermal degradation. In both thermal and catalytic experiments, the activity was strongly enhanced by an increase in the waste oil/tetralin ratio, which shows the latter is not the best solvent for the liquefaction of aliphatic plastics. The catalytic hydrocracking of HDPE and PP over Zr02 modified by SOfand WOZ- groups and promoted by Pt and Ni as hydrogenation metals has been investigated by Venkatesh et al.3 HDPE was almost completely converted over Ni- or Pt-containing Zr02-SO4 catalysts at 375C and 82 atm of cold-

MWP / None

MWP I FHYD / Zn

MWP / ZSM-5

Milk Jugs / ZSM-5


I
I

Coke Bottles I ZSM-5

/ 20
I

40

60

/ 80
1

/
100

Oil Yield (%)


f diflerent Figure 6.1 Oil yields obtained in the thermal and catalytic hydrogenation o P H= ~ 54 atm, plastic wastes (T = 430 "C, t = 60 min, starting solvent = tetralin)

.'

Hydrogenat ion

163

Oil Yield(%)
loo

1
................
....

................

....

....

....

Thermal

FHYD/Si/Al

ZSM-5

Waste Oil(%)
00% a 3 3 %a 6 7 % mIOO%

f the waste automotive oilltetralin ratio on the oil yield in the thermal Figure 6 . 2 Efect o and catalytic hydrogenation of a commingled plastic waste (T = 445 "C, t = 60 min, starting PH2= 54 atm).2 (Reprinted from Coal Science, eds. J.A. Pajares and J.M.D. Tascon. 01995, pp. 15 19-1 522, with permission from Elsevier Science)

hydrogen pressure after just 25 min of reaction. A high yield of gasoline products was obtained with high ratios of iso/normal alkanes. Likewise, PP could be hydrocracked with high conversions and at a lower temperature (325C) than in the case of HDPE, the products obtained being richer in branched alkanes due to the presence of alternate side methyl groups in the PP backbone. Similar results were obtained over the Ni- and Pt-containing Zr02-W03 catalysts, although they exhibited a higher stability than the sulfated materials, which undergo a loss of sulfur at temperatures above 250 "C. Solid superacids, based on sulfated Zr02 and Fe203, were also used by Zmierczak et aL4 for the degradation of PS under hydrogen atmospheres. The products obtained at temperatures between 350 and 450 "C consisted mainly of aromatic hydrocarbons such as benzene, ethylbenzene, diphenylalkanes and minor quantities of cyclization products, polyphenyls and triphenylalkanes. Only trace amounts of styrene were detected in both the thermal and catalytic experiments, which is a consequence of the combined effects of hydrogen and the catalysts. Increasing the temperature or the initial hydrogen pressure (34170 atm) leads to an increase in the yield of fully depolymerized products, mainly benzene and alkylbenzenes. The catalyst having the highest acidity, sulfated Zr02, also showed the highest PS conversion, favouring both depolymerization and dealkylation reactions. The authors propose a mechanism to explain the formation of the main products directly from the PS chains. The initial step involves the generation of carbocations either by hydride abstraction

164

Chapter 6

at benzylic positions or by protonation of aromatic rings, which are followed by a- and p-cleavage reactions. have investigated the catalytic hydrocracking of HDPE and a Ding et sample of commingled post-consumer plastic, consisting mainly of HDPE and small amounts of PP, PS and other polymers. The degradation experiments were conducted at 375 "C with a cold-hydrogen pressure of about 68 atm. The catalysts were based on Ni and NiMo sulfides loaded on a hybrid support, formed by a mixture of ZSM-5 zeolite and amorphous silica-alumina. Under these conditions, virtually no thermal hydrocracking occurred, whereas the different catalysts tested led to HDPE conversions in the range 65-80%. Increasing the catalyst loading from 20 to 40% allowed almost total HDPE conversion to be achieved. The products obtained were about 45% gases and 55% liquids in the gasoline range. The latter were mainly branched and linear paraffins, some aromatics and cycloparaffins, with few olefins. Figure 6.3 shows the various hydrocarbons detected by GC-MS analysis of the liquid products. The high iso/normal alkane ratio observed in both gases and liquids was related to hydroisomerization reactions, also promoted by the metal sulfides present in the catalysts. Similar results were obtained in the hydrocracking of plastic wastes, high conversions being achieved with both the Ni- and NiMo-containing catalysts. These results were in contrast to the behaviour of ZSM-5 alone, which showed lower activity for the conversion of the post-consumer plastic mixture compared to the HDPE degradation. The authors concluded that the Ni- and NiMo-based catalysts were more resistant than purely acidic solids to deactivation by N- and S-containing compounds present in the raw plastic waste (0.65 wt% N and 0.0 1 wt% S). The hydrodenitrogenation and hydrodesulfurization activities of the Ni and NiMo catalysts were confirmed by the fact that N and S were not detected in the produced oils. In terms of both conversion and oil yield, the best results in the degradation of the plastic waste were obtained with the Nicontaining catalyst, and no synergistic effects were observed by the incorporation of Mo into the catalyst. Likewise, the authors studied the effect of an increase in the initial hydrogen pressure in the range 17-68 atm, which improved both conversion and liquid yield. Moreover, some changes were also found in the composition of the oil fraction, with a reduction in the content of aromatics and olefins and an increase in the cycloparaffin proportion as the initial hydrogen pressure was increased. Thus, at 375 "C and a cold-hydrogen pressure of 68 atm, the liquid obtained consisted of 7.5% paraffins, 47.1% isoparaffins, 20.6% cycloparaffins, 1.7% olefins and 23.170 aromatics. Compared with a commercial premium gasoline, this oil is of better quality because it contains more isoparaffins and less aromatics. Catalytic hydrogenation of PE, PP, PS and PET has also been investigated by Rothenberger et a1.6 over NiMo/A1203and zeolite 13X catalysts. The reactions were carried out in autoclaves at 430 "C and 68 atm of cold-hydrogen pressure. Incorporation of 1-methylnaphthalene as solvent had a negative effect on the HDPE conversion, regardless of catalyst or atmosphere used. It is concluded that the solvent may favour retrograde polymerization reactions. PE and PP

2000000

18OOOOfl

1600000

1400000

12C1000Q

1000000

I-

d
I

800000

600000

400000

200000

4.00

Figure 6.3 GC-MS analysis of the liquid products obtained in the HDPE hydrocracking over a catalyst consisting of Ni supported on amorphous silica-alumina and ZSM-5 (80/20%).5Reaction conditions: T = 375 "C, t = 1 h, starting P H 2 = 68 atm. (Reprinted with permission from W. Ding, J. Liang and L.L. Anderson, Energy Fuels, 11, 1219. 0 1997 ACS)

c3

u ch l

166

Chapter 6

were converted mainly into aliphatic hydrocarbons whereas PS was depolymerized into alkylbenzenes. HDPE was the most difficult of the model polymers to convert. PET degradation leads to the formation of gases with significant quantities of both carbon dioxide and ethane, whereas a substantial amount of terephthalic acid was detected in the liquid products. The authors propose that PET degradation takes place by a sequential mechanism, consisting of the initial cleavage of C-0 bonds with liberation of ethylene, which is hydrogenated into ethane, followed by cleavage or reduction of the remaining carboxyl groups. The lower reactivity of PE compared with PS and PET under hydrocracking catalytic conditions (430 "C, 86 atm of cold-hydrogen pressure) has also been observed by Joo and Curtis7 over a presulfided NiMo/A1203 catalyst. Thus, after 1 h of reaction, about 94% of PS and PET were converted, whereas the reaction with LDPE yielded a conversion of only 72%. The thermal and catalytic degradation of a sample of commingled plastics (CP) under hydrogen atmosphere has been studied by Ibrahim and Seehra8 through ESR experiments in order to detect the formation of radicals at different temperatures. The starting plastic waste contained about 95% HDPE, 5% PP and 3% PET. Figure 6.4 shows the ESR spectra of the CP sample alone and mixed with 10% sulfur. At room temperature, the sample gave no ESR signal, indicating the total absence of free radicals in the starting material. As the sample is heated, the first hint of an ESR signal appears near 360"C, which is very close to the threshold temperature of the HDPE thermal decomposition. However, when the CP sample was mixed with 10% sulfur, the first ESR signals were clearly observed at temperatures as low as 285"C, showing an increased formation of radicals in the presence of sulfur. Further decreases in the threshold temperature were observed with CP and 10% S in a hydrogen atmosphere (270 "C) or when CP was mixed with 10% of S + NiMo/

CP

CP 4- 1O%S
244'C
298%

Figure 6.4

ESR spectra of a commingled plastic sample recorded at diferent temperatures, alone and with 10% sulfur.8 (Reprinted with permission from M.M. Ibrahim and M.S. Seehra, Energy Fuels, 1997,11,926.0 1997 ACS)

-+ +38 1'C --.LcL.rccch

3 6 0 ' C

4 0 7 ' C

396C

Hydrogenation

167

A1203 catalyst (240C). On the contrary, the incorporation of ZSM-5 zeolite into the CP sample did not cause a decrease in this temperature. The effects of S, NiMo/A1203 and ZSM-5 zeolite on the degradation of commingled plastics is also depicted in Figure 6.5, which shows the changes in the free radical intensity, proportional to the free radical concentration, with temperature for mixtures of the CP sample with these materials. In the presence of NiMo/A1203 and ZSM-5 the free radical concentration is considerably reduced compared to the thermal degradation of CP, which indicates that they act as hydrogenation catalysts, capping the free radicals formed in the polymer. In contrast, a sharp increase in the intensity of the ESR signals is produced in the presence of sulfur. The authors propose that sulfur extracts hydrogen from the polymer structure, producing H2S and free radicals, which results in a lowering of the depolymerization temperature of at least 100 "C. However, the possible formation of undesired S-containing organic compounds was not investigated, results which would be essential if considering the possible applications of the produced oils. Two-step processes have also been proposed for the conversion of plastic wastes involving hydrogenation treatments. In this approach, hydrocracking is proposed for upgrading the oils produced by simple thermal decomposition of plastics. Thus, Joo and Guin' have investigated the hydrocracking of a residual gas oil fraction, produced by the thermal decomposition of a plastic mixture consisting of HDPE, PP and PS, followed by vacuum distillation at 90 "C.The reactions were conducted under 69 atm of cold-hydrogen pressure using a catalyst consisting of presulfided Ni-Mo supported on a zeolite-alumina mixture. Figure 6.6 illustrates the boiling point distributions of the liquids produced at 430 "C with different catalyst loadings and that corresponding to a

Figure 6.5

Free radical intensity (N) as a function o f the temperature for a commingled plastic sample, alone or mixed with sulfur and diferent catalysts.8 (Reprinted with permission from M.M. Ibrahim and M.S. Seehra, Energy Fuels, 1997, 11,926. 0 1997 ACS)

168

Chapter 6

aJ > .c .

40

20

0
0

100200300400500600

Boiling Point ("C)


Figure 6.6 Boilingpoint distributions of the liquids obtained by hydrocracking of aplastic pyrolysis oil at 430 "C with diferent catalyst loading^:^ without catalyst (a), 0.1 g loading (b), 0.7 g loading (c), 1 .O g loading (d). (Reprinted with permission from H.S. Joo and J.A. Guin, Energy Fuels, 11, 586. 0 1997 ACS)

commercial gasoline. Increased upgrading results from the higher catalyst concentration, leading to products that show a boiling point distribution very similar to the range of the commercial gasoline.

3 Hydrocracking of Rubber and Used Tyres


The most commonly used rubber in tyre manufacture is styrene-butadiene copolymer containing about 25 wt% of styrene. The presence of a high concentration of double bonds in the rubber backbone makes the alternative of degrading rubber wastes by treatment in hydrogen atmospheres very attractive. Moreover, because used tyres contain significant amounts of sulfur, hydrogenation also favours the removal of this undesired element as H2S, which allows oils to be produced with lower S content than those derived from tyre pyrolysis. Zmierczak et al.4 have investigated the catalytic hydrocracking of nonvulcanized rubber (SBR, styrene-butadiene copolymers) over superacid solids, consisting of sulfated Zr and Fe oxides. Figure 6.7 shows the GC-MS analysis of the liquids produced at 400C over sulfated Fe203, with assignments of the main peaks. Three types of product are observed: CS-Cg paraffins produced from the butadiene blocks of the polymer, alkylbenzenes derived from the

Hydrogenation

169

0.0

10.0

20.0

Retention time, min

Figure 6.7 GC-MS analysis of the liquid products obtained in the SBR liquefaction at 400 C over a suIfated Fez03 ~ a t a l y s t . ~ (Reprinted from Fuel Process Technol., 49, W. Zmierczak, X. Xiao and J. Shabtai, page 3 1. 0 1996, with permission from Elsevier Science)

styrene units, and bicyclic compounds, probably formed by catalytic side-chain cyclization reactions of alkylbenzenes. Increasing the temperature to 375450C with an initial hydrogen pressure of 102 atm leads to improvements in both conversion and gas yield, although the production of liquids remains with selectivities higher than 85%. Over sulfated zirconia, there are some changes in the product distribution, with a higher yield of ethylbenzene being obtained, which is related to the stronger acidity of the zirconia catalyst. Scheme 6.1 illustrates the mechanistic pathways proposed by Zmierczak et al.4 for the hydrocracking of SBR. According to the authors, the double bonds in the butadiene blocks may undergo partial hydrogenation and double bond migrations. P-Cleavage reactions of benzylic carbonium ions generate Cl-C3 alkylbenzenes and also longer alkyl and alkenylbenzenes with alkyl groups up to C4-Cl2. The latter can undergo intramolecular cyclization reactions to form cycloalkylbenzenes, indanes and tetrahydronaphthalenes. Likewise, /?-cleavage at both ends of the butadiene units may produce CS-Cl2 paraffins and olefins, as well as cycloparaffins by secondary cyclization. Finally, cycloparaffins can also be formed by dealkylation of cycloalkylbenzenes. A number of patents have been filed claiming the processing of waste rubber, either vulcanized or unvulcanized, by hydrocracking. The degradation usually proceeds in the presence of catalysts: CoNi/A1203,FeMo/Cr203, Mo acetates, CpZZrC12, etc. In some cases, a certain positive effect has been derived from the presence of sulfur compounds, which may either be the sulfur contained in the used tyre or H2S added to the reaction mixture. Liu et al.I4 have studied the non-catalytic hydrogenation of tyres in the

170

Chapter 6

(a) double bond migrations and pama1 hydrogenation


(b) Pxleavage reactions

(a) P-cleavage reactions at both ends of the

(b) partial hydrogenation and cyclization reactions

R = Cj - C3

R = C4 CIz alkyl

C5 - C,z paraffins and olefins C6 Clo cycloparaffins (mostly cyclopentanes and cyclohexanes)

and alkenyl
cyclization reactions (Iimired eaent)

t
dealkylalion

cycloalky Ibenzenes, indanes and tetralines

Scheme 6.1 Mechanistic pathways for the SBR depolyrnerization-liq~efaction.~ (Reprinted from Fuel Process Technol., 49, W. Zmierczak, X. Xiao and J. Shabtai, page 3 1. 0 1996, with permission from Elsevier Science)

temperature range 350400 "C.The conversion was dependent on the temperature and was improved by the addition of tetralin as solvent when working below 400C. However, at 400C the presence of tetralin or hydrogen has no effect on the conversion of the tyre. These results suggest that tyre liquefaction may follow a thermal cracking or a hydrogenation mechanism, depending upon the reaction temperature. At temperatures of about 350 "C, hydrogenation is a necessary step before bond cleavage whereas at 400 "C, thermal cracking becomes predominant. The maximum conversion obtained is about 66%, which is approximately the content of volatile matter in the raw tyre, suggesting that the carbon black in the tyre does not really react under these conditions. The total gas yield obtained at 400C was 6.2%, consisting basically of CO, C 0 2 and C l-Cs hydrocarbons. Anderson et a1.l' have studied the hydrocracking of scrap automotive tyres after removal of the steel thread and the textile netting present in the starting material. Figure 6.8 shows the results obtained in non-catalysed experiments carried out at 400 "C at different cold-hydrogen pressures. A maximum conversion into gas and oils of about 70% was obtained, confirming that the carbon black, which accounts for 31% of the tyre, is not reactive under these

Hydrogenation
U % G s N%Oils

171

7s

so

2s

0 0.1*

MPa in H2 (* in N2)

10

Figure 6.8 Product distribution obtained in the hydrogenation of a tyre at 400 "C as a function of the cold-hydrogen pressure.'5 (Reprinted with permission from L.L. Anderson, M. Callin, W. Ding, J. Liang, A.M. Mastral, M.C. Mayoral and R. Murillo, Ind. Eng. Chem. Res., 36,4763. 0 1997 ACS)

conditions. The oil/gas ratio was dependent on the hydrogen pressure; the higher the pressure the lower the gas yield. Thus, only small amounts of gases were produced at cold-hydrogen pressures above 49 atm. GC-MS analysis of the oils showed they contain alkylbenzenes, alkylnaphthalenes and long-chain hydrocarbons in the range C8-C20. When the degradation was carried out in the presence of an Fe catalyst (red mud), the production of gases was minimized at all the pressures studied. The composition of the liquids was very similar to the non-catalysed reactions. The major difference was the presence of large amounts of 1-methyl-4-isopropylcyclohexane, which indicates that a greater extent of hydrogenation occurs in the catalysed tyre degradation.

Coliquefaction of Coal and Plastics

In recent years a number of studies have addressed the coprocessing of plastic waste and coal in a hydrogen atmosphere. Interest in this alternative is due to the higher H/C ratio present in most plastics compared to coal, the latter being hydrogen deficient in regards to the targeted liquid products. Accordingly, during coliquefaction of coal and plastic mixtures it is expected that the polymers act as hydrogen donors to the coal. There are two main benefits to this approach: plastic wastes are removed by feedstock recycling while the consumption of hydrogen gas necessary for coal liquefaction is reduced. Currently, the transformation of coal into liquids is not economically competitive due to the cost of hydrogen, However, as described below, some contradictory data have been reported in the literature regarding whether there are synergistic effects in the coprocessing of coal and plastics. Taghiei et al.' have studied the coliquefaction of bituminous and sub-

172

Chapter 6

bituminous coals with PE and a mixed plastic waste in the presence of several catalysts: ion exchanged iron, ultrafine ferrihydrates and ZSM-5 zeolite. The oil yield of the mixture was improved by about 10% compared to the processing of coal and plastics individually in both thermal and catalytic tests, which was attributed to the presence of synergistic effects. The authors conclude that the plastic acts as a hydrogen donor solvent for the coal. In a further work, Huffman et a1.2 compared the coprocessing of coal with PE and PP at 400C and 54 atm of cold-hydrogen pressure over ZSM-5 zeolite, concluding that PP undergoes more synergistic reactions with coal than PE. Anderson et a1.16 have investigated the coliquefaction of a high volatile bituminous coal with a commingled waste polymer mixture, containing mostly HDPE. Figure 6.9 shows the conversion and oil yield obtained over a variety of catalysts in comparison with the non-catalysed treatment. The best catalyst, in terms of both activity and production of oils, consisted of Pt supported on amorphous silica-alumina. However, in all cases the conversions of the coalplastic mixtures were lower than those expected from the results obtained in the processing of coal and plastics individually. In a further work,17 the authors explored a two-step process based on the hydrogenation of a mixture of coal and the liquid fraction produced in the catalytic cracking of HDPE at 435 "C under nitrogen. The increase in conversion obtained showed that the liquid products from HDPE degradation are more compatible with coal than the starting polymer. Coliquefaction of coal and model plastics (PE, PP, PS and PET) in two- and

5% Ni on SiO21A203
2%Pt on SiO21A12Q SiO2IAl203 (25%) Thermal Run
0

I
10

20
%

30

40

50

moil Yield

OConversion %

Figure 6.9 Thermal and catalytic coliquefaction of a coal+ HDPE mixture (.50/.50%). I 6 Reaction conditions; T = 430C, t = 1 h, starting PH2 = 68 atm, 10 wt% catalyst. (Reprinted from Coal Science, eds. J.A. Pajares and J.M.D. Tascon. 0 1995, pp. 15 15-1 5 17, with permission from Elsevier Science)

Hydrogenation

173

multicomponent mixtures has also been studied by Rothenberger et aL6 Although the conversion of mixtures was higher than that of the individual components, the authors related this improvement to artefacts of the analytical procedure. They concluded that no synergism is observed in the coliquefaction of coal and plastics, although the presence of the latter does not hinder the coal conversion provided the amount of plastics is limited. Likewise, Joo and Curtis7 have studied the catalytic coprocessing of plastics with coal and a petroleum residue over an NiMo/A1203 catalyst at 430 "C and 86 atm of cold-hydrogen pressure. Binary mixtures of the petroleum residue with PS and PET resulted in high conversions (about %YO), whereas the coprocessing of a residue-LDPE mixture led to 72% conversion. Lower activities were observed in the coliquefaction of coal and plastic mixtures, the lowest conversion corresponding to the system coal-LDPE. Moreover, introduction of plastics into the coal-residue system caused a decrease in both conversion and formation of hexane-soluble products. The possible presence of synergistic effects was evaluated from the calculation of coprocessing effect factors, taking into account the results obtained with the individual components and mixtures of them. Binary systems including plastics, especially LDPE, resulted in negative factors, showing the absence of synergistic effects and the low compatibility of plastics with coal.

5 Coliquefaction of Coal with Rubber and Used Tyres


Coprocessing of coal with rubber wastes has the same potential advantages as mentioned earlier for coal and plastic mixtures: removal of these wastes by conversion into fuels and reduction in the amount of hydrogen consumed in the coal conversion. In fact, coliquefaction of coal and used tyres was studied before the conversion of coal-plastic mixtures, also leading to better results regarding the presence of synergistic effects during the coprocessing of these two materials. In addition to the role of rubber as a hydrogen donor, the carbon black present in the used tyre has been found to have a positive catalytic effect on the conversion of coal. The following advantages were described by Farcasiu et a1.'8-20 when investigating the coprocessing of coal and used tyres: the products obtained were of better quality than those produced in the thermal conversion of coal with tetralin as solvent; the ratio of rubber to coal could be varied over a wide range (0.74-1.5) without significantly changing the yield and quality of the products; and the presence of carbon black increased both the conversion and oil yield. In a typical experiment at 425 "C with 136 atm of initial hydrogen pressure, over 80% of the products were found in the heptane-soluble fraction with a high hydrogen content (about 9.7%) and low heteroatom content (1.7% oxygen, 0.47% sulfur and 0.4% nitrogen). The authors propose that this liquid could be a good replacement for the petroleum-derived aromatic oil used in tyre manufacture. Liu et a1.I4 also observed a synergistic effect in the liquefaction of coal and

174

Chapter 6

tyre mixtures, because the conversions obtained were higher than those of the individual components. The effect of the tyre was similar to that of tetralin, although the latter was somewhat more effective in promoting coal conversion. In a later work,21the differences between the conversion of mixtures of coal with tetralin and tyres were reduced in the presence of an iron sulfide catalyst. While the authors did not observe any catalytic effect of the carbon black, they explained the tyre effect by the generation of small aliphatic free radicals by the rubber component, which cap coal-based radicals, resulting in a decrease in retrogressive reactions and, therefore, in higher coal conversions. The effect of the different components present in used tyres on the generation of free radicals on coal has been studied by Ibrahim and Seehra22through ESR experiments. Figure 6.10 shows that the addition of carbon black or SBR to coal at 420 "C leads to a significant increase in the free radical intensity. Likewise, it was observed that the tyre tread and its components lower the temperature of coal thermal cracking, favouring the formation of free radicals on the coal. Mastral et al.23924 have also investigated the effect of the main components present in tyres (carbon black, styrene-butadiene copolymer and polybutadiene) on the liquefaction of coal. Coprocessing of coal and carbon black confirmed the catalytic role of the latter, as it promotes hydrocracking reactions leading mainly to the formation of gaseous products. The addition of SBR to coal improves the yield of gases, oil and asphalt fractions, even at relatively low temperatures (350-375 "C).It is proposed that SBR favours the stabilization of the radicals involved in the process through alkylation reactions

0 ,

2
c

1.4

0
TY

./ := +4
styrene butadiene(3546) added

''

carbon(35%) added
1 I ,

time (rnin.)

Figure 6.10 Free radical intensity (N) as a function of time at 420 "Cfor coal alone, and after the addition of carbon black and styrene-butadiene copolymer.22 (Reprinted from Fuel Process Technol., 45, M.M. Ibrahim and M.S. Seehra, page 213. 0 1995, with permission from Elsevier Science)

Hydrogenation

175

of the aromatic rings present in this elastomer. Finally, the role of polybutadiene at low temperature is of little significance, whereas at 400 "C its effect on the coal conversion is similar to that of SBR, probably due to the aromatization of the aliphatic radicals generated from polybutadiene under these conditions. Anderson et af.15 have studied the effect of hydrogen pressure on the coprocessing of coal and scrap automotive tyres at 400 "C. While no synergism was detected in a nitrogen atmosphere, coliquefaction of coal and tyres under hydrogen resulted in enhanced conversions and yields of asphalt and oil fractions, particularly at low hydrogen pressures. It is proposed that the radicals coming from the tyre cannot break down the coal structure, although they may stabilize radicals formed by hydrogenation reactions. The effect of the composition of both the coal and the waste tyres during their coliquefaction has been investigated by Tang and under both thermal and catalytic conditions. Bituminous coals yielded higher conversions than lower rank lignite and sub-bituminous coals, while waste tyres having typical contents of rubber, carbon black, aromatic oil and ZnO proved to be good coal solvents. Addition of slurry phase hydrotreating catalysts, Mo and Fe naphthenates, and sulfur improved the coal conversion. On the contrary, little effect was observed by incorporation of carbon black under these conditions. In a further work, Tang and Curtis26have studied the influence of the tyre components by replacing coal with model compounds: 4 4 1-naphthylmethyl)bi benzy 1 (NMBB), dibenzo t hiop hene and 5-met hy 1- 8-( 1-met hy let hy1)dibenzothiophen-4-01 (MMDH). Carbon black was active for the NMBB hydrocracking, whereas little activity was found with SBR, waste tyres and waste tyre liquefaction residues. However, when the latter residues were previously heat treated to remove organic coatings and recover the carbon black component, significant NMBB conversions were observed. Moreover, both carbon black and heat-treated residues were active for the hydrodesulfurization of dibenzothiophene and MMDH, particularly when combined with Mo naphthenate and S. These results confirm the catalytic properties of carbon black in coal-tyre coliquefaction, provided that its surface area is accessible. A somewhat different alternative for the coprocessing of coal and tyres is the application of two-stage processes. In this case, tyres are first pyrolysed to yield an oil which is then used as solvent in the coal liquefaction. With this approach, the problem associated with the separation of finely divided carbon black dispersed in the viscous oil produced from coal can be avoided. Harrison and Ross27,28 have studied coal liquefaction using a tyre pyrolysis oil (TPO) as a cosolvent in addition to a hydrogenated anthracene oil. TPO was separated by distillation into two fractions having boiling points above and below 275 "C (TPO+ and TPO-, respectively). GC-MS analysis of these oils showed that they are mainly composed of aromatic hydrocarbons. Liquefaction experiments on coal were carried out in the temperature range 380-420 "C with mixtures of TPO or TPO+ and the anthracene oil. Increasing the proportion of TPO or TPO+ led to a decrease in the extent of dissolution, although the reduction was less than that expected from the measured decrease in the

176

Chapter 6

hydrogen donor content of the solvent. When coal liquefaction was performed with prehydrogenated TPO, the extent of dissolution was even higher than that produced with the anthracene oil alone. These results indicate that TPO, especially after hydrogenation, presents good hydrogen shuttling characteristics. Likewise, Orr et al.29930 have explored the possible use of tyre pyrolysis oil as a solvent for coal liquefaction. The potential of this alternative was demonstrated by the fact that coal-TPO mixtures were transformed with higher conversion than when coal was reacted directly with ground waste rubber tyres. It is proposed that the polyaromatic compounds present in the TPO favour coal dissolution during liquefaction. Treatment of coal-TPO mixtures (50/50%) at 430 "C under 68 atm of cold-hydrogen pressure in the presence of a M o catalyst led to a high coal conversion in just 10 min of reaction. From electron probe microanalysis of the coal particles after the reaction, the authors conclude that TPO favours the catalyst dispersion and its contact with coal, which results in enhanced coal conversion.

6 Summary
Hydrocracking of plastic and rubber wastes has been investigated in recent years as a method of degrading these polymers by conversion into liquid and gaseous products. The reactions are carried out under pressure and in many cases in the presence of solvents. Compared to non-hydrogenating treatments, conversion of plastics and rubber in a hydrogen atmosphere leads to higher proportions of liquid products, which are formed mainly by paraffinic and aromatic hydrocarbons. Moreover, hydrogenation is expected to favour the removal of the heteroatoms present in the starting wastes in the form of volatile compounds (HCl, H2S, etc.). Most of the studies on plastic and rubber hydrogenation are based on the use of catalysts to promote cracking and hydrogenation reactions. A wide variety of catalysts have been examined, consisting typically of transition metal compounds (Fe, Mo, Ni, Co, etc.) supported on an acid porous matrix (alumina, amorphous silica-alumina, zeolites, sulfated zirconia, etc.). A certain reduction in the conversion of plastics such as HDPE, LDPE and PP has been observed in both thermal and catalytic hydrocracking compared with the corresponding treatments in an inert atmosphere. Likewise, the addition of hydrogen-donating solvents to the reaction mixture is not beneficial, also leading to a decrease in plastic conversion. Hydrogen atoms probably react with radicals and carbocations generated from the polymer by thermal and/or catalytic cracking to form saturated products, which hinders the progress of the degradation. In the liquefaction of used tyres, the maximum conversion usually obtained is about 70%, indicating that only the rubber component is converted and the carbon black remains unreacted during hydrogenation. Increasing the pressure allows the oil yield to be improved, the production of gases being almost

Hydrogenation

177

completely suppressed. The oils produced consist mainly of alkylbenzenes, alkylnaphthalenes and bicyclic compounds. However, little information can be found on the sulfur content of these oils. The possibility of coprocessing plastic or rubber wastes with coal has been extensively studied in recent years. Interest in these processes is due to the possible role of the polymers as hydrogen donor agents to the coal. In this way, synergistic effects have been observed in the coliquefaction of used tyres and coal with enhanced oil production compared to the conversion of the individual components. However, there is some controversy over the coprocessing of coal and plastic mixtures because, although earlier works claimed the existence of synergistic effects, several authors have not found any improvement in the coliquefaction of these two materials. Better results have been obtained when coprocessing coal and plastic pyrolysis oils, because the latter are more compatible with coal than the starting polymer.

7 References
1 M.M. Taghiei, Z. Feng, F.E. Huggins, and G.P. Huffman, Energy Fuels, 1994, 8, 1228. 2 G.P. Huffman, Z. Feng, F.E. Huggins, and Y. Mahajan, in Coal Science, ed. J.A. Pajares and J.M.D. Tascon, Elsevier Science, Amsterdam, 1995, p. 1519. 3 K.R. Venkatesh, J. Hu, W. Wang, G.D. Holder, J.W. Tierney, and I. Wender, Energy Fuels, 1996, 10, 1163. 4 W. Zmierczak, X. Xiao, and J. Shabtai, Fuel Process. Technol., 1996,49, 31. 5 W. Ding, J. Liang, and L.L. Anderson, Energy Fuels, 1997, 11, 1219. 6 K.S. Rothenberger, A.Y. Cugini, R.L. Thompson, and M.V. Ciocco, Energy Fuels, 1997, 11, 849. 7 H.K. Joo and C.W. Curtis, Energy Fuels, 1996, 10,603. 8 M.M. Ibrahim and M.S. Seehra, Energy Fuels, 1997,11,926. 9 H. S. Joo and J.A. Guin, Energy Fuels, 1997, 11, 586. 10 J. Alpert, US Patent 3 704 108, 1972. 11 M. Morita and T. Takamatsu, US Patent 4251 500,1981. 12 P.R. Stapp, US Patent 5 158 983, 1992. 13 C.J. Gibler, L.R. Chamberlain, R.A. Kemp, and S.E. Wilson, US Patent 5 162446, 1992. 14 Z . Liu, J.W. Zondlo, and D.B. Dadyburjor. Energy Fuels, 1994,8,607. 15 L.L. Anderson, M. Callh, W. Ding, J. Liang, A.M. Mastral, M.C. Mayoral, and R. Murillo, Ind. Eng. Chem Res., 1997,36,4763 16 L.L. Anderson, W. Tuntawiroon, and W.B. Ding, in Coal Science, ed. J.A. Pajares and J.M.D. Tascon, Elsevier Science, Amsterdam 1995, p. 1515. 17 W.B. Ding, W. Tuntawiroon, J. Liang, and L.L. Anderson, Fuel Process. Technol., 1996,49,49. 18 M. Farcasiu, C.M. Smith, E.P. Ladner, and A.P. Sylvester, Fuel Chem. Preprints, Am. Chem. Soc., 1991,36, 1869. 19 M. Farcasiu and C.M. Smith, US Patent 5061 363, 1991. 20 M. Farcasiu, Chemtech, 1993,23, 22. 21 Z. Liu, J.W. Zondlo, and D.B. Dadyburjor, Energy Fuels, 1995,9,673. 22 M.M. Ibrahim and M. S. Seehra, Fuel Process. Technol., 1995,45, 213. 23 A.M. Mastral, R. Murillo, M.J. Perez-Surio, and M. Callen, Energy Fuels, 1996, 10, 941.

178

Chapter 6

24 A.M. Mastral, R.Murillo, M. Callen, M.J. Pkrez-Surio, and M.C. Mayoral, Energy Fuels, 1997, 11, 676. 25 Y. Tang and C.W. Curtis, Fuel Process. Technol., 1996,46, 195. 26 Y. Tang and C.W. Curtis, Energy Fuels, 1997,11, 1143. 27 G. Harrison and A.B. Ross, in Coal Science, ed. J.A. Pajares and J.M.D. Tascon, Elsevier Science, Amsterdam, 1995, p. 1531. 28 G. Harrison and A.B. Ross, Fuel, 1996,75(8), 1009. 29 E.C. Orr, J.A. Burghard, W. Tuntawiroon, L.L. Anderson, and E.M. Eyring, Fuel Process. Technol., 1996,47,245. 30 E.C. Orr, Y. Shi, Q. Li, L. Shao, M. Villanueva, and E.M. Eyring, Energy Fuels, 1996, 10, 573.

CHAPTER 7

Concluding Remarks
Plastics and rubber are essential materials in todays industrialized societies. The consumption of plastics has grown by a factor of about 60 in the past 30 years, which has led to a corresponding increase in the generation of plastic wastes. One of the most valuable properties of plastics, their low density, is one of the major limitations in the recycling of plastic wastes. Thus, to recover one tonne of plastics it is necessary to collect about 20000 plastic bottles. Plastic wastes are mainly found in municipal solid wastes (MSW). As a consequence of their low density, plastics account for just 8 wt% of the MSW, but this value increases to over 20% in volume terms. In spite of the great diversity of plastic materials, plastic wastes are made up of a relatively small number of polymers: polyethylene, polypropylene, polystyrene, polyvinyl chloride and polyethylene terephthalate. These resins account for more than 90% of total plastic wastes. Of the various alternatives for plastic waste management (landfilling, mechanical recycling, energy recovery and feedstock recycling) the conversion of plastic and rubber wastes into valuable chemicals and fuels is probably the most interesting approach for the future. Landfill sites are becoming progressively more scarce, and the cost of burying waste is also increasing. Incineration is an effective treatment for the removal of solid wastes, which enables the generation of electric energy from the combustion heat. However, incineration is socially rejected in many countries because of the possible emission of noxious compounds into the atmosphere. Moreover, incineration of plastics and rubber does not allow the chemical value contained in these polymeric materials to be recovered. Likewise, mechanical recycling of plastics is an alternative limited by the contamination and degradation that most polymers suffer during their use. Goods made of recycled resins are often of lower quality than those manufactured from virgin polymers, only being useful for undemanding applications. Bearing in mind these factors, chemical or feedstock recycling seems to be the ideal solution for plastic waste management. However, only a very small amount of plastic wastes is currently subjected to feedstock recycling, because most of the existing processes are not economically feasible. A similar picture is valid for the management of rubber wastes, the major destination being also landfilling and incineration.
179

180

Chapter 7

Five main methods of feedstock recycling have been considered in this book, classified according to the degradation conditions and the products obtained: chemical depolymerization, gasification, thermal treatments, catalytic cracking and reforming, and hydrogenation. Chemical depolymerization by reaction with agents such as water, glycols, alcohols, ammonia, etc. allows the polymer chains to be broken down to yield the starting monomers, which in turn can be used to synthesize fresh polymers with properties similar to the virgin resins. Chemolysis methods have been applied to a variety of polymers: polyesters, polyamides, polyacetals, polyurethanes and polycarbonates. Several commercial processes are currently in operation based on this alternative. However, chemical depolymerization is of no use for the recycling of most addition polymers, which are the main components of the plastic waste stream. Another limiting factor is the high degree of purity required in many polymerization processes, which makes it necessary to carry out intensive separation and purification steps of the degradation products. Therefore, the development of combined processes may be of interest in order to benefit from the advantages of the individual treatments. Thus, PET degradation by glycolysis-hydrolysis is a process which combines the efficiency and higher rate of glycolysis with the production by subsequent hydrolysis of high purity terephthalic acid, which can be easily repolymerized into PET. Gasification of plastic and rubber wastes proceeds by reaction with air, oxygen or steam at temperatures in the range 700-1600C. The obtained synthesis gas is a highly versatile product which can be used in a number of chemical syntheses leading to ammonia, methanol, hydrocarbons, etc. An important advantage of this process is that it requires little or no pretreatment and separation of the waste stream, which makes the coprocessing of plastics with other solid wastes possible. However, the produced syngas must be cleaned to remove various contaminants and pollutants, which requires a number of operations. Another limiting factor of gasification is that to be profitable it must be used in large capacity plants. In contrast with the information available on the conversion of mixed solid wastes, a lack of fundamental research exists on the gasification of the various plastic and rubber polymers. In addition to gasification, other partial oxidation processes have been studied for the degradation of plastic and rubber wastes, such as reaction with organic peroxides or treatment with oxygen under supercritical water conditions. Likewise, polyolefin thermooxidation with oxygen at low temperature has been proposed to yield wax oxidates with commercial applications. However, most of the studies on the oxidative degradation of polymers have been aimed at promoting the polymer stability rather than favouring its decomposition into valuable products. Decomposition of plastic and rubber wastes by treatment at temperatures between 400 and 800 "C in inert atmospheres has been extensively investigated. A wide variety of processes and reactors have been developed for the thermal conversion of plastic and rubber materials. Sand fluidized bed reactors are one of the preferred systems as they favour heat and mass transfer during the

Concluding Remarks

181

polymer decomposition. Only in the thermal degradation of a few polymers (PS, PMMA), is the starting monomer selectively recovered; in most cases the products consist of complex mixtures of hydrocarbons. Conversion of plastics at temperatures above 600C typically yields a high proportion of gases rich in olefins, and minor amounts of oils with a high concentration of aromatic hydrocarbons. The olefin yield is increased when the thermal degradation is carried out in the presence of steam. On the other hand, low temperature processes lead mainly to oils formed by a mixture of n-paraffins and a-olefins distributed over a wide range of carbon atom numbers. The possible application of these oils as fuels is limited by the high concentration of olefins, which makes their further upgrading in refinery units necessary. However, the presence of C1 hinders their refinery processing because there is usually a C1 limit of 10 ppm. Accordingly, a key issue that must be resolved is how to completely remove the C1 remaining in the oils. Thermal treatment of plastic mixtures containing significant amounts of PVC is especially problematic. At temperatures between 250 and 300 "C, most of the C1 evolves as HC1, which leads to a corrosive environment. Moreover, interaction between the HCl released and other components in the wastes has been reported, causing undesired secondary reactions. In recent years, new interesting alternatives have been proposed for the thermal conversion of plastic wastes, such as their coprocessing with coal or degradation in the presence of solvents, although more intensive study of these methods is needed. Thermal conversion of rubber wastes has mainly addressed the degradation of used tyres, which yields three fractions: gases, oils and a solid residue. The gases are proposed in most cases to be used as fuel gas, while the oils are rich in aromatic hydrocarbons. The presence of around 1 % of sulfur in the oils limits their direct application as fuels, hence it is recommended they are upgraded and desulfurized in refinery units. However, the possible application of these oils as a source of valuable chemicals has not been explored. The char fraction is obtained with yields of around 30%, being made up of the carbon black initially present in the tyres, together with significant amounts of sulfur and zinc. The conversion of pyrolytic chars into activated carbons through various treatments has been investigated as a method of obtaining higher added-value products, which may improve the profitability of the tyre recycling process. The quality of the products derived from the decomposition of plastics can be increased by the use of different catalytic systems. Catalytic cracking has been mainly investigated for the conversion of PE, PP and PS. The catalyst promotes polymer cleavage, and allows the reaction to proceed under milder conditions, and selective formation of the desired fractions. By selection of a suitable catalyst, it is possible to obtain different products: olefinic gases, gasoline fractions and middle distillates. The liquid fractions are valuable fuels, because they contain mainly branched alkanes and aromatics, with properties similar to commercial gasolines. Catalysts typically used include acid solids such as amorphous silica-alumina and zeolites, although some interesting results have also been obtained with basic catalysts and activated carbons. The major

182

Chapter 7

drawback of these processes is the pretreatment of the plastic waste stream required to remove all those components (additives, fillers, C1- and N-containing polymers, contaminants, etc.) that may damage the catalyst activity. Although it is assumed that C1-containing polymers may cause a poisoning of the catalyst, very few studies have been performed on the effect that the presence of PVC may have on the catalytic degradation of polyolefins. Studies on the catalytic degradation of real plastic wastes would be preferable to working with model plastics. On the other hand, the high viscosity and low heat transfer of the molten polymers hinder their contact and conversion over the catalyst in conventional fixed bed reactors. These problems have been solved in part through the dissolution of plastics in liquid petroleum fractions or by a combination of thermal and catalytic treatments. In this case, the catalyst reforms the products obtained by thermal degradation of the polymer, which implies that the rate of the whole process is controlled by the thermal step. Only a few works have been reported on the use of more complex but also more effective reactors such as catalytic fluidized beds. New reactor types should be developed to promote the direct catalytic cracking of plastic wastes. Hydrogenation, usually in the presence of catalysts, is the final method of feedstock recycling considered. Degradation of plastic and rubber wastes at temperatures around 400 "C and different hydrogen pressures allows the oil yield to be increased and highly saturated products to be obtained. The absence of olefins in the oils enables their use as fuels without further upgrading. Moreover, hydrogenation favours the removal of the heteroatoms present in the raw wastes as HC1, H2S, NH3, etc. Catalysts commonly used in these processes are bifunctional solids having both cracking and hydrogenationdehydrogenation activity. However, the economic feasibility of hydrogenation is questionable, owing to the cost of hydrogen and the high pressures required. Coliquefaction of plastic and rubber wastes with coal has recently been investigated. Although the benefits of coprocessing plastics and coal is the subject of controversy, some synergism has been observed in the coliquefaction of used tyres and coal. Forecasts about the future of plastic waste management place feedstock recycling as one of the faster growing alternatives. However, although many of the feedstock recycling processes for plastic and rubber wastes were developed some time ago, only a few of them have reached a commercial stage. In most cases, the major limitation is the economy of the process. The profitability of any recycling process depends on a number of factors: purity of the raw waste, capital investment in the processing facilities, value of the products obtained, etc. In many cases, these factors are interdependent. Thus, the quality of the oils derived from thermal degradation of plastics is related to their C1 content, which is a function of both the treatment conditions and the PVC concentration in the raw wastes. Progress in plastic separation and sorting is a key factor in the future development of both mechanical and feedstock recycling. As methods of plastic sorting are improved, the purity and homogeneity of the plastic wastes is increased, which favours the application of more advanced degradation

Concluding Remarks

183

processes, specific for each type of plastic. Therefore, more intensive work on plastic separation should be carried out in the next few years. Special interest should be devoted to the development of methods for the identification and separation of PVC, because its presence, even at low concentrations in the plastic waste mixture, leads to a number of difficulties in most feedstock recycling alternatives. Moreover, the problems associated with plastic sorting and further conversion should be simultaneously addressed, because in many cases the most suitable separation procedure may depend on the type of transformation subsequently applied. According to the pretreatment requirements the following order of feedstock recycling methods can be envisaged: gasification < thermal treatments z hydrogenation < catalytic cracking < chemical depolymerization. Increasing the value of the products derived from feedstock recycling of polymeric wastes will also help the process economy. Conventionally, degradation of plastic and rubber wastes has led to complex mixtures of hydrocarbons, suggested for use as fuels, in many cases of low quality. However, little effort has been focused on using the products resulting from polymer degradation in chemical synthesis, with the exception of the various chemolysis methods which lead to the starting monomers. In recent years, several studies have addressed the production of higher added value chemicals: gaseous olefins by high temperature pyrolysis or through catalytic treatments, waxes and a-olefins by thermal degradation of polyolefins, activated carbons from the chars obtained in used tyre decomposition, indane compounds from tyre pyrolysis oils, etc. As a final conclusion it can be stated that, although feedstock recycling of plastic and rubber wastes has been studied for more than 20 years, due to the need for suitable alternatives for dealing with the increasing generation of these types of wastes and the number of technological challenges still present, this topic will continue to be the subject of much research activity for the beginning of the next century.

Subject Index
Accelerators, 12 Activated carbon as cracking catalyst, 133 for PE catalytic cracking, 142 for PP catalytic cracking, 148 from tyres gasification, 66 from tyres pyrolysis, 121 Activators, 13 Addition polymers, 7 Adipic acid as polyesterification agent, 36 from nylon-6,6 hydrolysis, 53 Adsorbents, use of chars as, 122 Alumina catalysts, 130 for catalytic hydrocracking of mixed plastics, 164 for catalytic hydrocracking of rubber, 169 for coliquefaction of coal + plastic + petroleum residues, 173 for PE catalytic cracking, 134, 143 for plastic wastes hydrogenation, 161 for PP catalytic cracking, 145 for PP catalytic cracking, 148 for PP + nylon-6 catalytic cracking, 151 Aminolysis of PET, 41 of PU, 49 Ammonia from synthesis gas, 62 reaction with PET, 41 reaction with PU, 49 reaction with polyamides, 53 Ammonolysis of PET, 41 PU, 49 Antistatic agents, 12 Antioxidants, 11, 13 effect on the PVC thermal degradation, 96 Antiozonants, 13 Aromatic hydrocarbons from PE catalytic cracking, 143 from PS catalytic hydrocracking, 163 ASR, see Automotive shredder residue Automobile rear lights, thermal depolymerization, 99 Automotive shredder residue thermal degradation, 104 catalytic cracking, 150 Basic oxide catalysts, for PS catalytic cracking, 152 Basic salt catalysts, for catalytic cracking of mixed plastics, 150 Benzene, from PS catalytic cracking, 148, 152 Beta, zeolite, 133 for PP catalytic cracking, 147 BHET, see Bis(hydroxyethy1)terephthalate Bifunctional catalysts, 142, 161 Bisfenol A, from polycarbonate hydrolysis, 55 Bis(hydroxyethyl)terephthalate, from PET, 32, 33,42 Blowing agents, 12 Boudouard reaction, 6 1 BP process, 111 Carbon black, 13, 173, 174 E-Caprolactam from nylon-6 hydrolysis, 53 from nylon-6 thermal degradation, 100

185

186 L-Caprolactone, as polycondensation agent, 36 Catalytic cracking, 129 alumina catalysts, 130 bifunctional catalysts, 142 catalysts, 130 catalyst deactivation, 138, 142, 146, 149 catalyst regeneration, 142 combined with thermal cracking, 151 Friedel-Crafts catalysts, 130, 133 of mixed plastics, 150 of plastic wastes, 129 of polyethylene, 133 of polypropylene, 145 of polypropylene + nylon-6 mixtures, 151 of polystyrene, 148 of rubber wastes, 150 silica-alumina catalysts, 131 zeolite catalysts, 131 CaX, zeolite, for PP catalytic cracking, 146 Celluloid, 1 Char from PE pyrolysis, 83 from tyre pyrolysis, 118 use as adsorbent, 122 Chemical depolymerization, 3 1 Chemolysis, 3 1 comparison of methods for PET, 44 Coal coliquefaction with plastics, 171 coliquefaction with plastic + petroleum residues, 173 coliquefaction with rubber, 173 coliquefaction with tyres, 173 coprocessing with plastic wastes, 1 15 coprocessing with lignocellulosic and plastic wastes, 116 Coke deposition on catalysts, 139, 143, 146, 155 from PE pyrolysis, 83 from tyres pyrolysis, 120 Coliquefaction of coal + plastic mixtures, 171 of coal + rubber mixtures, 173 of coal + tyre mixtures, 171 of coal + tyre pyrolysis oils, 176 Colorants, 11 Condensation polymers, 8

Subject Index

Continuous kinetic model for thermal degradation, 80 Copolymer, 4 Coprocessing of lignocellulosic and plastics wastes + coal, 116 of plastics wastes + coal mixtures, 115 Deactivation of catalysts, 138, 142, 146, 149,164 Degradative extrusion, 65, 113 Dehydrochlorination, of PVC, 9 1 Diamines, from PU hydrolysis, 48 Diethylene glycol, reaction with PET, 36 Dimethyl carbonate, from polycarbonate chemolysis, 56 Dimethyl terephthalate, from PET, 32,37, 44 Dioxines, 17 from PVC gasification, 64 from PVC thermal degradation, 94 DMT, see Dimethyl terephthalate Elastomers, 4, 6 , 7 Engineering plastics, 1 EPS, see Expanded polystyrene Ethyl benzene from PS catalytic cracking, 148, 152 from PS thermal degradation, 88 Ethylene, from PE ultrapyrolysis, 83 Ethylene glycol from PET hydrolysis, 38 from PET methanolysis, 37 reaction with PET, 33,42 reaction with PU, 46 Ethylene oxide, reaction with PET, 36 Expanded polystyrene, 10 FCC, see Fluidized catalytic cracking FCC catalysts, for PS catalytic cracking, 149 Fibres, 4 Fillers, 11, 13 Fisher-Tropsch process, 62 Fixed bed reactor, 81, 83,86,88, 100, 101, 118, 136, 138, 142, 144, 146, 153, 156 Flame retardants, 11 Fluidized bed gasifier, 67 internally circulating reactor, 83, 108

Subject Index
reactor, 80, 81, 86,91, 93,98, 105, 111, 144, 149, 156 Fluidized catalytic cracking of plastics + gas oil mixtures, 145 of plastics + rubber wastes, 150 Ford hydroglycolysis process, 50 Formaldehyde, from polyacetals hydrolysis, 54 Friedel-Crafts catalysts, 130 for PE conversion, 133 Fuji Recycle-Mobil Oil process, 156 Furans, from PVC thermal degradation, 94 Gaseous hydrocarbons from catalytic hydrocracking of mixed plastics, 164 from PE catalytic cracking, 140, 145 from PE pyrolysis, 80 from PP thermal degradation, 86 from reforming of PE pyrolytic oils, 153 Gasification, 59 coupled with thermal degradation, 65 of plastic wastes, 62 of plastic wastes and carbonaceous solids mixtures, 66 of plastic wastes and heavy oils mixtures, 65 of polyethylene, 62, 66 of polypropylene, 62,66 of polystyrene, 66 of PVC, 64 of rubber wastes, 62 of used tyres, 66 reactions, 60 Gasoline from catalytic hydrocracking of mixed plastics, 164, 167 from PE catalytic cracking, 137, 141, 145 from PE catalytic hydrocracking, 163 from PP catalytic cracking, 146 from reforming of PE pyrolytic oils, 153, 155 G1ycolysis of PET, 33 of PET + PVC mixtures, 36 of PU + polyurea mixtures, 47 HDPE, see High density polyethylene

187 High density polyethylene, 8 High impact polystyrene, 10 HIPS, see High impact polystyrene HY, zeolite for PE catalytic cracking, 134 for PP catalytic cracking, 145 for PS catalytic cracking, 148 Hydrochloric acid, from PVC, 75,91, 101, 108,111,114 Hydrocracking catalysts deactivation, 164 of coal + plastic mixtures, 171 of coal + rubber mixtures, 173 of coal + tyre mixtures, 171 of mixed plastics, 164 of PE, 162 of plastic wastes, 161 of PP, 162 of PS, 163 of rubber, 168 of tyres, 168 Hydrodenitrogenation, 164 Hydrodesulfuration, 164, 175 Hydrogen sulfide, from rubber and tyres hydrocracking, 168 Hydrogenation, 161 bifunctional catalysts, 161 Hydrolysis of PET, 38 of PET + polyolefins mixtures, 41 of polyacetals, 54 of polycarbonates, 54 of PU, 47 of PU + rubber mixtures, 48 i-PP, see Isotactic polypropylene Isotactic polypropylene, 9 KFS-16, for PE catalytic cracking, 141 Krupp Uhde PreCom gasification process, 67 L, zeolite, for PP catalytic cracking, 146 LDPE, see Low density polyethylene Lewis acid catalysts, in polyamide ammonolysis, 54 Light stabilizers, 11 Linear low density polyethylene, 9 Lithium acetate, 33

188 LLDPE, see Linear low density polyethylene Low density polyethylene, 9 Maleic anhydride, as polyesterification agent, 35 MCM-41 for PE catalytic cracking, 140 for PP catalytic cracking, 146 properties, 132 Metal acetates, 34 Metal oxides, effect on PVC thermal degradation, 96 Methanation reactions, 61 Methanol from synthesis gas, 62 reaction with PET, 37 supercritical, 38 superheated vapor, 38 Methanol ysis of PET, 37 of PET with supercritical methanol, 38 of PET with superheated methanol, 38 Methyl methacrylate, from PMMA depolymerization, 98 Middle distillates for PE catalytic cracking, 141 for PP catalytic cracking, 146 Mixed plastics automated sorting, 25 catalytic cracking, 150 dissolution, 23 flotation, 22 hydrogenation, 162 hydrocracking, 164 manual separation, 22 PET + polyolefin hydrolysis, 41 PET + PVC glycolysis, 36 PU + polyurea glycolysis, 47 PU + rubber mixtures hydrolysis, 48 sorting and separation, 22, spectroscopic identification, 25 thermal degradation, 100 Mixed solid wastes, gasification, 66 Molten salt catalysts, 150 Monomer, 4 Mordenite, 133 for PE catalytic cracking, 138 for PP catalytic cracking, 146 for PS catalytic cracking, 149

Subject Index
MSW, see Municipal Solid Wastes Municipal Solid Wastes, 14 composition, 14 Natural rubber, 1, 12 N a y , zeolite, for PE conversion, 137 Nylon-6, 52 acid hydrolysis, 52 ammonolysis, 53 Nylon-6,6, 52 ammonolysis, 53 basic hydrolysis, 53 Oils from PE pyrolysis, 8 1 from PP thermal degradation, 86 from tyres pyrolysis, 119 Olefins from PE catalytic cracking, 140 from reforming of PE pyrolytic oils, 153 Oligomer, 4 Organic peroxides, 69 Oxidation catalysts, 69 Oxidative degradation, 70 0 x 0 synthesis, 62 PAH, see Polyaromatic hydrocarbons Parak process, 1 13 Partial oxidation of plastics wastes, 69 of rubber, 69 PET, see Polyethylene terephthalate Plastics additives, 11,93 consumption, 2,6 uses, 2 Plastic wastes, 13 catalytic cracking, 129 chemical recycling, see Feedstock recycling composition, 14 energy recovery, 17 environmental impact, 15 feedstock recycling, 20 hydrocracking, 161 hydrogenation, 161 gasification, 59,62 generation, 13 management, 16,27 mechanical recycling, 19

Subject Index

189 glycolysis-methanolysis, 43 hydrolysis, 38 methanolysis, 37 methanolysis-hydrolysis, 42 neutral hydrolysis, 40 non-aqueous alkaline degradation, 39 thermal degradation, 100 Polyisocyanurates from PET glycolysis, 34 glycolysis, 47 glycolysis-aminolysis, 52 Polymerization, mechanism, 7 Polymers classification, 4 dissociation energy, 75 Polymethyl methacrylate thermal depolymerization, 98 thermal depolymerization in the presence of solvents, 99 Pol yols from PET glycolysis, 34 from polyisocyanurates glycolysis, 48 from PU aminolysis, 49 from PU aminolysis-hydrolysis, 52 from PU ammonolysis, 49 from PU glycolysis, 46 from PU hydrolysis, 48 Poly(p-methylstyrene), thermal degradation, 89 Polypropylene, 9 catalytic cracking, 145 catalytic hydrocracking, 162 gasification, 62,66 oxidative degradation, 70 pyrolysis, 86 thermal degradation, 85 Polystyrene, 10 catalytic cracking, 148 catalytic hydrocracking, 163 combined thermal and catalytic cracking, 152 degradation in the presence of water, 90 gasification of, 66 oxidative degradation, 70 pyrolysis, 90 thermal degradation, 86 thermal degradation in the presence of solvents, 89 Poly(styrene-ally1 alcohol), thermal degradation, 89

partial oxidation, 59 recovery, 14 recycling, 17 recycling policy, 27 reduction, 16 refinery processing, 21,23, 114 reuse, 16 tertiary recycling, see Feedstock recycling Plastic wastes and heavy oils mixtures gasification, 65 hydrogenation, 162 Plasticizers, 1 1, 13 effect on the PVC thermal degradation, 95 PMMA, see Polymethyl methacrylate Pollutants in the synthesis gas, 61 Polyacetals, hydrolysis, 54 Polyamides, 52 hydrolysis, 52 pyrolysis, 100 Polyaromatic hydrocarbons, 88 Polycarbonates, hydrolysis, 54 Polycondensation, 33 Polyenes, from PVC, 91 Polyesterification, 35 Polyesterification agents, 35 Polyesters chemical depolymerization, 32 pyrolysis, 100 Pol yet hylene catalytic conversion, 133 catalytic hydrocracking, 162 combined thermal and catalytic cracking, 151, 153 gasification, 62,66 steam cracking, 83 thermal degradation, 78 types, 8 Polyethylene terephthalate, I0 acid hydrolysis, 38 alkaline hydrolysis, 39 aminolysis, 41 ammonolysis, 41 chemical depolymerization, 32 comparison of chemolytic methods, 44 glycolysis, 33 glycolysis kinetics, 34 glycolysis mechanism, 33 gl ycolysis-hydrolysis, 42

190

Subject Index

Polytetrafluorethylene, gasification, 66 Polyurethanes, 45 aminolysis, 49 aminolysis-hydrolysis, 52 ammonolysis, 49 combined chemolysis methods, 50 Ford hydroglycolysis process, 50 from PET glycolysis, 34 glycolysis, 46 glycolysis-aminolysis, 52 hydrolysis, 47 pyrolysis, 100 Polyvinylchloride, 10 dioxines from thermal degradation of, 94 furans from thermal degradation of, 94 gasification, 64,66 oxidative degradation, 70 thermal degradation, 91 thermal degradation mechanism, 75 types, 10 PP, see Polypropylene Primary amines, reaction with PET, 41 Process oils, 13 Propylene, from PP pyrolysis, 86 Propylene glycol, reaction with PET, 35 Propylene oxide, reaction with PET, 36 PTFE, see Polytetrafluorethylene PS, see Polystyrene PU, see Polyurethanes PVC, see Polyvinylchloride Pyrolysis, 74 of polyamides, 100 of polyesters, 100 of polyethylene, 80 of polypropylene, 86 of polyurethanes, 100 of polystyrene, 88 of tyres, 118

REY, zeolite for PE catalytic cracking, 134 for PP catalytic cracking, 145 for PS catalytic cracking, 148 for reforming of PE pyrolytic oils, 153 Rotary kiln reactor, 111 Rotating cone reactor, 82, 86, 110 Rubber, 12 additives, 12 catalytic cracking, 150 degradation with supercritical water, 122 gasification, 62 hydrocracking, 168 hydrocracking mechanism, 169 hydrogenation, 161 oxidative degradation, 71 thermal degradation, I I7 thermal degradation in the presence of solvents, 122 SBR, see Styrene-butadiene rubber Shape selectivity, 131 Silica-alumina catalysts, 131 for catalytic hydrocracking of mixed plastics, 164 for coliquefaction of coal and PE, 172 for PE catalytic cracking, 134, 137, 138, 140, 141, 143, 151 for plastic wastes hydrogenation, 161 for PP catalytic cracking, 145, 146 for PS catalytic cracking, 148, 152 for reforming of PE pyrolytic oils, 153 Sodium carbonate, from polycarbonates hydrolysis, 55 Sodium hydroxyde, reaction with PET, 39 s-PP, see Syndiotactic polypropylene Steam cracking, of PE, 83 Styrene from PS catalytic cracking, 152 from PS thermal degradation, 88 from PS pyrolysis, 90 Styrene-butadiene rubber, 12 Sulfated zirconia, 133 for PE catalytic cracking, 140 for PE catalytic hydrocracking, 162 for plastic wastes hydrogenation, 161 for PP catalytic hydrocracking, 162 for PS catalytic hydrocracking, 163 for rubber catalytic hydrocracking, 168

RCD model for thermal degradation, 79 Reducing gas, from synthesis gas, 62 Reforming of thermal cracking products, 129 Regeneration of catalysts, 142 Reinforcements, 11 Repeating units, 7 Resin, 4 Retarders, 13

Subject Index

191 Two-zone reactor, 15 1, 152 Tyres catalytic cracking, 150 composition, 1 17 gasification, 66 hydrocracking, 168 non-catalytic hydrogenation, 169 thermal degradation, 117 UHMWPE, see Ultrahigh molecular weight polyethylene U1tra high molecular weight pol yet hy lene, 9 Unsaturated polyesters from PET glycolysis, 34 glycolysis, 36 USY, zeolite for PE catalytic cracking, 138 for PP catalytic cracking, 147 for reforming of PE pyrolytic oils, 153 Veba Oil process, 111 Vulcanization, 12 Water-shift reaction, 60 Waxes from PE thermal cracking, 84 from plastics thermal degradation, 1 I3 Wax oxidates, 70 Winkler gasification process, 67

Sulfur, for catalytic hydrocracking of mixed plastics, 164 Sulfuric acid, reaction with PET, 38 Supercritical water oxidation, 70 water degradation, 122 methanol, 38 Syndiotactic polypropylene, 9 Synthesis gas from carbonaceous materials, 59 uses, 61 Teflon, see Polytetrafluorethylene Terephthalic acid, from PET, 32,38,42 Thermal cracking, 74 combined with catalytic cracking, 15I Thermal degradation, 73 continuous kinetic model, 80 in the presence of solvents, 89, 113, 122 mechanisms of, 74 of ASR, 104 of PE + PP + PS + PVC + PET mixtures, 101, 108 of PET, 100 of plastic mixtures, 100 of polyethylene, 78 of poly(p-methylstyrene), 89 of poly(styrene-ally1 alcohol), 89 of PP, 85 of PS, 86 of PVC, 91 of PVC + aluminium mixtures, 103 of rubber, 117 of tyres, 117 RCD model, 79 Thermal depolymerization of PMMA, 98 of PMMA from automobile rear lights, 99 Thermal stabilizers, 11 Thermoplastics, 5, 7, 8 Thermoselect gasification process, 68 Thermosets, 5, 7, 11 Toluene, from PS thermal degradation, 88 TPA, see Terephthalic acid TPA amide, from PET, 41 TPS Termiska gasification process, 67 Transesterification catalysts, 33, 37,40 Trioxane, from polyacetals hydrolysis, 54

X, zeolite, for hydrocracking of mixed plastics, 164


Y , zeolite, 132

Zeolites, 131 X, 164 Y, 132 ZSM-5, 132 Ziegler-Na tta catalysts, 9 Zinc acetate, 33,41 Zinc oxide, 13 ZSM-5 zeolite, 132 for catalytic hydrocracking of mixed plastics, 164 for catalytic reforming of mixed plastic thermal degradation products, 156 for coliquefaction of coal and plastics, 172

192

Subject Index

ZSM-5 zeolite (cont.) for hydrogenation of mixed plastics,


162

for PP catalytic cracking, 146 from PS catalytic cracking, 149, 152 for reforming of PE pyrolytic oils, 153

for PE catalytic cracking, 135, 138, 140,


141, 144, 151