Recycling of PVC and Mixed Plastics Wastes by F. P. La Mantia

Francesco Paolo La Mantia, Editor
Recycling of PVC
and
Mixed Plastic Waste
0CP
ChemTec Publishing
Copyright © 1996 by ChemTec Publishing
ISBN 1-895198-11-9
All rights reserved. No part of this publication may be reproduced, stored or transmitted in any form
or by any means without written permission of copyright owner. No responsibility is assumed by the
Author and the Publisher for any injury or/and damage to persons or properties as a matter of products
liability, negligence, use, or operation of any methods, product ideas, or instructions published or
suggested in this book.
Printed in Canada
ChemTec Publishing
38 Earswick Drive
Toronto-Scarborough
Ontario M1E 1C6
Canada
Canadian Cataloguing in Publication Data
La Mantia, Francesco Paolo

Recycling of PVC and mixed plastic waste
Includes bibliographical references and index.

ISBN 1-895198-11-9
1. Polyvinyl chloride--Recycling. 2. Plastic scrap--

Recycling. I. Title
TP1122.L35 1996 668.4’236 C95-920960-3

iii
TABLE OF CONTENTS
Recycling of PVC: Effect of the Processing Operation 1
Gerald Scott
1 Polymers in Waste Management 1
1.1 Photo-biodegradation of Plastics Litter 2
1.2 Composting of Plastics in Mixed Household and Garden Waste 2
1.3 Materials Recycling of Segregated, Clean Industrial Waste 3
1.4 Chemical Recovery from Mixed Plastics Waste 3
1.5 Incineration with Heat Recovery 4
1.6 Sanitary Landfill 4
2 Polymer Processing: A Damaging Operation 4
3 Mechanodegradation of PVC 5
4 The Effects of Plasticisers, Impact Modifyers and “Compatibilisers”
on PVC Performance 7
4.1 Plasticisers 7
4.2 Impact Modifiers 8
4.3. “Compatibiliser” (Solid Phase Dispersants) 9
5 Stabilisation of PVC 10
5.1 HCl Scavenging 10
5.2 Removal of Conjugated Unsaturation 11
5.3 Removal of Monoenic Unsaturation 11
5.4 Hydroperoxide Decomposition 13
5.5 Chain-breaking Antioxidants 13
5.6 Reagents for Allylic Chlorine 14
6 Antagonism and Synergism 15
7 Antioxidant-modification of Polymers During Processing 16
8 Conclusions 18
Acknowledgment 19
References 19
Investigations on the Recycling Ability of Used PVC 23

D. Braun and K. Krämer
Introduction 23
Characterization of PVC Waste 25
Measurement of Residual Stability 26
Analysis of the Heat Stabilizer and Other Additives 28
iv
Determination of Molecular Weight 31

Possibilities of Additional Stabilization of Used PVC 32
Addition of Heat Stabilizers 33
Improvement of Thermal Stability by Fillers 33
Results of Processing Experiments with Filled PVC 34
Practical Examples of PVC Recycling 37
Conclusions 40
References 41
Recycling of Complete PVC Windows 43

Herbert Uhlen
Introduction 43
Dismantling System 44
Conclusions 48
Recycling PVC Bottles and Pipes by Coextrusion 51

Georges Voituron
Introduction 51
Mechanical Recycling 53
General Approaches 53
Regeneration of PVC Pipe 53
Regeneration of PVC Bottles 54
Re-use of PVC Bottles and Pipes for Coextruded Pipes 57
Description of the Coextrusion Process 57
Formulation of the PVC 57
Thermal Stability 57
Lubrication 59
Impact Resistance 59
Expandability and Foaming Process 59
Flow Stability 60
Modification of the Production Equipment 60
Product Evaluation 61
Conclusions 61
References 62
v
Basic Concepts on the Recycling of Homogeneous and

Heterogeneous Plastics 63
Francesco Paolo La Mantia
Introduction 63
Main Problems in Plastics Recycling 64
Degradation 64
Influence of Structure and Morphology on the Behavior of Polymers 66
Molecular Weight 67
Branching 69
Polymer Blends 70
Conclusions 75
Acknowledgment 76
References 76
Compatibilization of Recycled Polyethyleneterephthalate/

Polypropylene Blends Using a Functionalized Rubber 77
B. Ballauri, M. Trabuio, and F. P. La Mantia
Introduction 78
Experimental 78
Materials 78
Recycling 79
Structural Determinations 79
Rheological and Mechanical Properties 79
Results and Discussion 80
Structural Characterization 80
Rheological Properties and Processability 85
Mechanical Properties 88
Conclusions 91
References 91
Recycling of Post-consumer Plastic Containers for Liquids 93

Introduction 93
Experimental 95
Materials 95
Processing 97
vi
Mechanical Properties 97
PVC Determination 97
Results and Discussion 98
Effect of Stabilizers and Lubricants 98
Fillers 102
Modifying Agents 105
Compatibilizers 107
Comparison Between Different Systems 109
Conclusions 114
Acknowledgment 116
References 116
A New Process for Recycling of Mixed Plastics Waste 117

B. Dubrulle D’Orhcel
Introduction 117
The Newplast-process Technology 118
Description of a Newplast-process Installation 120
Homomicronization Equipment 122
Newplast-process 124
Characterization of the Mixed Recycled Polymers 125
Conclusions 127
References 127
Grafted Copolymers and Their Use in Polymer Blends 129

B. Boutevin, Y. Pietrasanta, and Jean-Jacques Robin
1 Introduction 129
2 The Major Methods of Synthesis of Grafted Copolymers 131
2.1 The Combination Reaction 131
2.2 Macroradicals 132
2.2.1 Degradation of Polymers 132
2.2.1.1 Mechanical Degradation 132
2.2.1.2 Ultrasonic Degradation 133
2.2.2 Chemical Route 133
2.2.2.1 Oxidation-reduction 133
2.2.2.2 Diazotation 133
2.2.3 Irradiation 133
vii
2.3 Transfer to Polymer 134

2.4 Macroinitiators 134
2.4.1 Oxidation 134
2.4.2 Ozonization 134
2.4.3 Preirradiation in the Presence of Air 135
2.5 Macromonomers 135
2.5.1 Radical Polymerization 135
2.5.2 Polyaddition Reaction 136
3 Synthesis of Grafted Copolymers by Ozonization of Polymer 136
3.1 Mechanism of Ozonization 136
3.2 Determination of the Polymer Activation 138
3.2.1 Colorimetric Titration 138
3.2.2 Selective Titration of Hydroperoxides in the Presence of
Ferrous Salts 138
3.2.3 Iodometric Titration 139
3.3 Determination of the Dissociation Constant of Peroxy Functions 139
3.3.1 Determination of the Global Dissociation Constant Kd 139
3.3.2 Determination of Kd Constant and the Efficiency factor, f 140
3.3.3 Comparison Between the Dissociation Constants 141
3.4 The Grafting Methods 142
3.4.1 Grafting in Solution 142
3.4.2 Grafting in Bulk 142
3.5 Analysis of Products 142
3.6 Examples of Applications of Grafted Copolymers 146
3.6.1 Particular Case of Plastics Recycling 146
4 Polymer Blends and Alloys 147
4.1 PE/PVC Blends 148
4.2 PE/PS Blends 149
4.3 PS/PVC Blends 154
4.4 PS/PVC/PE Blends 155
4.5 PVDF/PS Blends 157
4.6 EVA/PS Blends 157
5 Monomaterial Waste Sources Application to Polymer-Bitumen
Blends 157
5.1 Polyethylene Wastes 157
5.2 Crosslinked Ethylene-vinyl Acetate, EVA, Copolymer Wastes 162
viii
6 Conclusions 163
References 163
Effect of Contamination on the Recycling of Polymers 167

J. Scheirs and G. Camino
Introduction 167
Contamination by Other Polymers 169
PVC in PET 169
HDPE in PET 169
PP in HDPE 170
PET in HDPE 170
Cellulose in HDPE 170
Multilayer Polymers 170
Contamination by Packaging Contents 171
Contamination by the Environment 173
Soil in Mulch Film 173
Moisture Contamination in PET 173
Contamination by Additives and Their Conversion Products 173
Chemical Modification of the Polymer 176
Contamination During Processin 177
Metal Contamination 177
Gel Contamination 178
Degraded Polymer Contamination 178
Detection of Contamination 178
Removal of Contamination 179
Removal of Contamination in Polyolefins 179
Removal of Contamination in PET 180
Removing Contamination in PVC 181
Removing Contamination in PC 181
Conclusions 181
References 182
UV Stability of Recycled Polymers 185

George Wypych
References 193
G. Scott 1
Recycling of PVC:
Effect of the Processing Operation
Gerald Scott
Aston University, Birmingham B4 7ET, UK
PVC presents particular difficulties in disposal due to the potential environmental hazards associated
with the chlorine content of the polymer. In principle materials recycling provides an ecologically ac-
ceptable way of re-utilising the energy content of the polymer but the processing operation is damag-
ing to the durability of recycled artifacts unless the provenance of the waste is known.
The principles involved in the protection of PVC against the mechanochemical damage that occurs
during processing are discussed for both rigid PVC and its modified forms. The importance of knowing
the previous history of PVC waste is emphasised, with particular reference to the stabilisers used. It is
suggested that the most effective method of recycling PVC is in a “closed loop”, so that the previous his-
tory and particularly the stabiliser formulation of the recovered polymer is known.
1 POLYMERS IN WASTE MANAGEMENT

There is no unique solution to the problem of plastics waste and litter. A hierar-
chy of complementary conservation and disposal techniques are required to
solve the escalating problem of waste management in industrial societies.
It is generally agreed that the “systems” approach to waste management
1
should include the procedures outlined below. There is as yet no universal
agreement as to the order of importance in which they should be placed. It is the
opinion of the present author that the preferred disposal technique must depend
on the type of waste and where it is located. For example, products that appear
mainly as litter must be considered quite separately from those which appear
predominantly in the sewage system and these should again be distinguished
from waste that appears in a controlled waste collection systems. Indeed, some
engineering polymers (e.g. from the automotive industry), should be considered
2 Recycling of PVC: effect of the processing operation
2
not as a waste but as a resource. In this case, closed loop recycling to the original
3
application with additional stabilisers is a realistic possibility.
1.1 PHOTO-BIODEGRADATION OF PLASTICS LITTER
The cost and energy usage in the collection and disposal of plastics which appear
as litter is unacceptable. Much of this arises from agriculture in the form of
mulching film, binder twine and increasingly irrigation tubing, silage bags, and
fertiliser packaging. Non-degradable plastics in agricultural land have an ad-
verse affect on the fertility of the soil and on the economics of automated agricul-
4
ture. Controlled photo-biodegradability is the only viable solution to this kind
of waste which is almost always heavily contaminated and thus unsuitable for
recycling even if it could be economically collected and segregated. By the same
token, the only way to reduce plastics packaging litter in the sea and on the
5
seashore is to ensure that it photodegrades and/or biodegrades rapidly. Some
polymers, notably those containing chlorine, present problems in
photo-biodegradation due to the potentially dangerous nature of the ultimate
low molecular weight degradation products, almost certainly containing chlo-
rine, that may be released into the environment. It has to be said, then that PVC
would not be the first choice for agricultural applications and this must be taken
into consideration at the materials design stage.
1.2 COMPOSTING OF PLASTICS IN MIXED HOUSEHOLD AND GARDEN WASTE
The technology of composting to give added value to domestic waste is likely to

increase rapidly during the next few years. It is claimed that by the end of the
century, every household in Germany will be connected to a municipal compost-
6
ing system and other European countries are expected to follow. Ironically, in
rural China and in other parts of the Far East, composting has always been the
most important way of using waste and with the advent of plastics, severe prob-
lems are now being experienced with non-degradable plastics packaging and
mulching film. Many manufacturers of degradable plastics are targeting the
6
composting environment for their products and several commercial photo-bio-
degradable polyolefins are readily oxidised and bioassimilated in aerobic
7
composters. For the reasons given above, biodegradable PVC would not be ac-
ceptable in such an application.
G. Scott 3
1.3 MATERIALS RECYCLING OF SEGREGATED, CLEAN INDUSTRIAL WASTE
Clean, segregated plastics arising from industrial operations (including the ulti-
mate disposal of motor car components) should be recycled in a closed loop sys-
3
tem into the primary application. Some packaging, notably film wrap, that can
be easily collected in bulk from industrial waste may be economically resourced
8
in the same way. Clean PVC waste can be effectively reprocessed in this way
provided the nature of the plasticiser/stabiliser formulation is known. This will
be discussed in detail in the following sections.
Mixed plastics packaging is best dealt with in one of the following ways.
1.4 CHEMICAL RECOVERY FROM MIXED PLASTICS WASTE
Some polymers, (e.g. polymethylmethacrylate, polystyrene) lead to good yields

9
of monomer on pyrolysis. Similarly, poly(ethylene terephthalate), PET results
in good recovery of monomers on hydrolysis. There is also a case for pyrolysing
contaminated mixed plastics packaging to obtain a mixture of hydrocarbons. It
10
can be seen from Table 1 that at moderate temperatures in a fluidised bed py-
rolysis unit, ethene and propene are major products with methane and ethane in
almost equal amounts. The latter and the scores of other compounds that have
been detected in small amount have fuel value.
Table 1: Composition of gaseous products from the pyrolysis

of mixed plastics, wt%10
Product 680* 735* 790*
Hydrogen 0.667 0.683 1.868
Carbon monoxide 8.376 14.155 6.316
Carbon dioxide 20.418 20.807 3.38
Methane 16.734 22.661 46.491
Ethene 18.383 20.69 25.994
Ethane 10.118 7.189 7.765
Propene 13.758 7.797 3.311
C3-C9 hydrocarbons 11.546 6.504 4.875
*temperature, oC
1.5 INCINERATION WITH HEAT RECOVERY
Many plastics have a higher calorific value than coal and in principle it would
seem logical to use waste plastics instead of fossil fuels. However, in Europe (al-
though not in Japan) public opinion is turning away from burning waste be-
cause of the perceived (although not necessarily real) dangers of toxic effluent.
PVC and the other chlorinated polymers present a particular problem in incin-
eration because of the massive evolution of hydrogen chloride. Nevertheless, it
can be readily accommodated in many modern incinerators by the use of calcium
oxide to absorb the hydrogen chloride if PVC is only a minor component of the
plastic waste. However, incineration is not seen by the public to be a safe process
in urban environments and as PVC is considered to be a major source of toxic ef-
fluent, PVC-containing wastes should be directed into recycling or landfill.
1.6 SANITARY LANDFILL
Disposing of plastics by burial in sanitary landfill, contrary to popular belief, is

by far the safest method of dealing with waste plastics, since, due to the absence
1,7
of oxygen, they do not oxidise or biodegrade under these conditions. Properly
stabilised PVC is stable almost indefinitely in anaerobic landfill and the chlo-
rine will remain locked away without harm to the environment almost indefi-
nitely. The main problem here is the shortage of landfill sites in many of the
developed countries. However, in contrast to the behaviour of putrescible mate-
rials, plastics do not cause subsequent subsidence and the land can be re-used.
2 POLYMER PROCESSING: A DAMAGING OPERATION
Organic polymers, unlike metals and glass, undergo irreversible changes in
chemical structure during manufacture and use. Thus, the metal content of a
can may be recovered in its entirety by a recycling process and, assuming it has
not been contaminated by other metals, can be refabricated to an identical can.
This is not possible with the commodity plastics, which undergo environ-
ment-induced oxidation during service which adversely affects their subsequent
performance.
Chemical changes begin in the extruder itself and unless steps are taken to
minimise damage by the use of additives, the process of screw extrusion has a
profoundly deleterious effect on the polymer structure and leads to the introduc-
tion of chemical impurities which affect the subsequent durability of the prod-
11
uct. Furthermore chemical and mechanical changes are accentuated during
G. Scott 5
recycling since the chemical defects introduced during the first processing oper-
ation and during subsequent use sensitise the polymer to further degrada-
8,12,13
tion. Stabilising systems have been developed to minimise degradative ef-
fects during the first processing operation (processing stabilisers) and during
use (heat and light stabilisers). Multicomponent stabiliser “package” are formu-
lated not only for individual polymers and their blends but more particularly for
specific applications. The constituent chemicals of most commercial antioxi-
dant/stabiliser packages are generally not disclosed and are almost always em-
pirical in origin.
However, the mechanochemistry involved in polymer processing is now
well understood and this provides a rational basis for design of antioxidants and
2
stabilisers for products which are to be recycled.
3 MECHANODEGRADATION OF PVC
Repeated recycling through the manufacturing process and exposure to the en-
vironment inflicts incremental chemical damage on the macromolecular struc-
ture of polymers and reduces the durability of fabricated products. Screw
extrusion involves the mechanical scission of the polymer chain in its viscous en-
vironment in the polymer melt. This produces highly reactive macroradicals at
the ends of the chain which, in the presence of the small amounts of oxygen dis-
solved in the polymer, cannot recombine but form peroxyl radicals and
11-18
hydroperoxides. This process, which is common to all polymers, is illustrated
16-18
for PVC in Scheme 1, reactions (a) and (b). However, in this case, other reac-
tions occur in addition to oxidation, of which the most important is loss of hydro-
gen chloride (Scheme 1, reactions (c) and (d)) to give unsaturation. This is the
major cause of discolouration in PVC due to “unzipping” of hydrogen chloride
from the polymer backbone. (Scheme 1). The two major products of
mechanodegradation, namely HCl and hydroperoxides (ROOH) react together
in a redox reaction (reaction 1) to give new and highly reactive oxygen radicals
and chlorine atoms which in turn attack the polymer chain to initiate further
elimination of HCI or initiate further oxidation in the PVC backbone (see
18
Scheme 2).
. .
PO2H + HCl → PO + H2O + Cl [1]
Scheme 1: Mechanooxidation of PVC18
Scheme 2: Sensitisation of PVC degradation by hydroperoxides
Alkoxyl radicals and chlorine atoms are highly reactive in hydrogen ab-
.
straction leading to the formation of macroalkyl radicals (P ), which in turn react
with almost zero activation energy with ground state oxygen which is itself a
diradical (reaction 2);
G. Scott 7
Figure 1. Relation between the initial rate of photooxidation of PVC and the concentration of func-
tional groups after processing. (Reproduced with permission from Developments in Polymer Sta-
bilisation-2, Ed. G. Scott, App. Sci. Pub. p. 61, 1980.)
. .
P + O2→ PO .2 → P + PO2H [2]
18
Photooxidation of PVC is initiated primarily by hydroperoxides. There is
a direct relationship between the hydroperoxide content of PVC after processing
and its rate of photooxidation as measured by carbonyl formation (see Figure 1).
Photolysis of in-chain hydroperoxides leads to rapid reduction in molecular
weight, with associated impairment of mechanical properties, notably elonga-
tion at break and impact resistance.
4 THE EFFECTS OF PLASTICISERS, IMPACT MODIFYERS
AND “COMPATIBILISERS” ON PVC PERFORMANCE
Modifying agents are added to PVC, sometimes in appreciable quantities in or-
der to improve performance in specific applications. Many of these have an ad-
verse effect on the stability of the polymer both during processing and on its
durability during subsequent exposure to the environment.
4.1 PLASTICISERS
Plasticisers have been used for very many years in order to reduce the glass
transition temperature of the polymer and give it rubbery properties at normal
19
temperatures. Plasticisers may have both positive and negative effects on PVC
during the processing operation. The advantage is that the temperature of the
processing operation may be substantially less than in the case of rigid PVC.
The disadvantage is that many plasticisers are readily oxidisable materials
forming hydroperoxides, which, as discussed in Section 3 lead to further oxida-
tion of the polymer substrate and reduction of mechanical properties. This is
particularly true when plasticised PVC is exposed to light and in this context,
the branched chain alkyl phthalate and phosphate esters are much more readily
19,20
photooxidised than the straight chain or aromatic esters. The use of conven-
tional chain-breaking antioxidant (e.g. hindered and semi-hindered phenols)
are essential for the protection of plasticisers from oxidation during mixing and
calendering processes. These operations are particularly damaging to the poly-
mer since the latter is readily accessible to oxygen of the atmosphere at the rela-
tively high temperatures involved. Recycling of plasticised PVC must therefore
be undertaken with considerable circumspection. It is not enough to simply add
new antioxidants and/or stabilisers to collected PVC waste. It is necessary to
know whether the mechanical properties and durability have been impaired
during first use. If they have it must be concluded that the stabilisation system
has been depleted or was not adequate in the first place. The cost involved in this
type of analytical investigation may be much more than the value of the product
made from the PVC waste.
4.2 IMPACT MODIFIERS
A major deficiency of PVC for many applications (for example in the building in-
21
dustry) is its poor resistance to impact. Consequently it became a common
practice during the 1970s to incorporate rubber-based modifiers such as ABS,
MRS, and MARS to improve the toughness of PVC. In this particular respect,
they were very successful and much higher levels of initial impact resistance
were achieved. However, PVC impact modifiers all contain butadiene segments
which are oxidatively very unstable, particularly at high temperatures and in
22-27 28,29
sunlight, so that their effect on the durability of PVC is catastrophic. Fig-
ure 2 shows that the incorporation of 10% of ABS into PVC causes a rapid decay
29
of impact strength during the first few hours of exposure in a weatherometer.
Degradation is initiated during the processing operation due to the formation of
24,25
hydroperoxides and although it can be retarded by the use of effective perox-
23
ide decomposing antioxidants, it cannot be eliminated. Recent research has
G. Scott 9
concentrated on eliminating the unsaturated components of the impact modifi-

ers, but it seems unlikely that rubber modified PVC will ever achieve the
thermooxidative and photooxidative stability of rigid PVC. For this reason, it is
not a good candidate for recycling to high quality products.
Figure 2. Effect of photooxidation on the falling weight impact resistance of PVC and impact modi-
fied PVC (10% ABS). £ unmodified PVC, , ∆ modified PVC (duplicate results). (Reproduced with
permission from European Polymer Journal, 13, 997 (1977)).
4.3. “COMPATIBILISER” (SOLID PHASE DISPERSANTS)
PVC often appears as a relatively minor component in mixed plastics waste in

which the major components are the polyolefins, particularly polyethylene.
When present in a blend with these polymers, it has a disastrous effect on me-
30
chanical performance. The addition of some rubbers (notably EPDM but not
the impact modifiers discussed in section 4.2) improves the impact strength of
30
PE/PVC blends very considerably, but relatively large amounts of these expen-
sive polymers are required and the unsaturation in the polymer again causes
problem with subsequent durability. The thermal and photooxidative stability
of polymer blends containing EPDM can be improved by the use of preventive
8
antioxidants (see Section 5). The reprocessing of blends of waste polymers con-
taining PVC as a substantial component is not at present a very promising ap-
proach. Segregation from the less dense polymers by flotation would seem to
offer the best prospect of success.
5 STABILISATION OF PVC
In most polymers (e.g. polyolefins) process stabilisation for first use can be
achieved with relatively low concentrations (0.05-0.1%) of hindered phenol or
phosphite antioxidant. The stabilisation of PVC requires much higher concen-
trations (generally 2-4% on the PVC content) in order to combat the mechani-
cally initiated HCl elimination referred to in Section 3. Some of them may also
stabilise the polymer to the environment, but their ability to do this depends en-
tirely on the amount of stabiliser that remains after the reprocessing operation
and how much chemical damage has been done to the polymer during the pri-
mary processing operation and in subsequent service.
It follows from the mechanism of mechanodegradation of PVC discussed
above that PVC may be stabilised in one or more of the following possible ways
18
during reprocessing.
5.1 HCl SCAVENGING
Since HCl is such a potent redox initiator for the decomposition of

hydroperoxides to radicals, its removal, as it is formed, constitutes an important
stabilisation mechanism. Most of the metal salts which form the basis of com-
mercial PVC stabilising systems act in this way. They include the long estab-
lished but now ecologically unacceptable, lead salts (e.g. lead carbonate), the
19
metal soaps (e.g. Ca, Zn, Cd, and Ba carboxylates). However, the most effective
class of PVC stabilisers encompasses the alkyl tin compounds and in particular
the alkyl tin maleates, I, and mercaptides, II. The main reason for the efficiency
of these agents is that, as well as being able to scavenge HCl (reactions 3 and 4),
they can also perform other functions as outlined below.
G. Scott 11
5.2 REMOVAL OF CONJUGATED UNSATURATION
It has long been recognised that maleic anhydride and its derivatives interrupt
the conjugated unsaturation in PVC and thus remove the colour due to this
20
mechanodegradation product, reaction 5.
By the incorporation of a maleate entity in a tin compound (I) complemen-

tary synergism occurs between the HCl scavenging function and the dienophilic
function. DBTM is not only a good processing stabiliser, it also protects the poly-
17
mer during UV exposure. However, it is partially or wholly removed during
16
processing and subsequent service. Figure 3 shows that during processing of
PVC in the presence of DBTM, the concentration of the latter is decreased
comensurately with the severity of the processing operation. Consequently, the
subsequent light stability of the product is compromised in the same proportion
(see Figure 4). The concentration of the stabiliser is similarly reduced on expo-
sure to the environment and when the polymer is reprocessed there may be little
or no stabiliser present. It is therefore essential that the stabiliser is replenished
before the polymer is reprocessed if it is to be used in durable applications, par-
ticularly out-of-doors.
5.3 REMOVAL OF MONOENIC UNSATURATION
As indicated above, isolated olefin groups are the first to be formed in PVC dur-
ing processing. The resulting allylic methylene groups are very readily oxidised
and their selective removal represents a very powerful preventive stabilisation
25,31
mechanism. Thiols are known to behave in this, way and it seems likely that
the heat stabilising effect of the dialkyl tin thioglycolate esters is due in part to
the liberation of thioglycolic esters, reaction 4, followed by their reaction with
32
monoenic unsaturation, reaction 6, and phosphite esters have been reported to
33
fulfill a similar function.
Figure 3. Decay of tin maleate i.r. carboxylate absorbance (1575 cm-1) during the processing of
PVC containing a synergistic mixture of dibutyl tin maleate (2.5 g/100 g) and Wax E (0.65 g/100 g.
− total sample; --- soluble phase. Processing temperatures are indicated on the curves. (Repro-
duced with permission from European Polymer Journal, 14, 913 (1978)).
Figure 4. Effect of processing time (at 210oC) on the photooxidation of PVC containing a tin
maleate stabiliser (numbers on curves indicate processing times in min.) (Reproduced with per-
mission from European Polymer Journal, 15, 51 (1979)).
G. Scott 13
ROCOCH2SH + CH2=CH- → ROCOCH2SCH2CH2- [6]
Reaction 6, where R contains a hindered phenol or a UV absorber function,

has also been used to introduce antioxidant and UV stabiliser groups into PVC
during processing (see Section 7).
5.4 HYDROPEROXIDE DECOMPOSITION
Compounds containing sulphide groups are widely used as synergistic antioxi-

34
dants in hydrocarbon polymers. They are the precursors of sulphur acids and
the dialkyl tin thioglycolates (II) have been shown to act in the same way in
35
PVC. The tin thioglycolates are much more effective heat stabilisers than the
tin maleates, almost certainly for this reason. Like the thiodipropionate esters,
they are sensitisers for the photodegradation of PVC due to the sensitivity of the
34
intermediate sulphoxides (e.g. III) to photolysis, giving initiating free radicals
(see Scheme 3). UV absorbers protect the sulphoxides from photolysis and they
show synergism with this class of light stabiliser.
Scheme 3: Photosensitizing action of the tin thioglycollates
5.5 CHAIN-BREAKING ANTIOXIDANTS
By themselves, chain-breaking donor antioxidants such as the hindered phe-

nols, IV, are ineffective processing stabilisers for PVC since alone they are not
able to inhibit the HCl unzipping reactions discussed above. However, they do
inhibit hydroperoxide formation and are widely used as synergists with the ba-
sic stabilisation systems discussed above and are always included in commer-
cial PVC stabiliser packages for plasticised formulations (see above,
Section 4.1).
5.6 REAGENTS FOR ALLYLIC CHLORINE
For many years it was believed that an important mechanism of PVC stabilisa-
tion was the replacement of labile chlorine atoms adjacent to double bonds in the
35
main chain. This conclusion was primarily based on the thermal stability of
model chlorine-containing compounds (e.g. V, VI) designed to simulate the
olefinic imperfections of the PVC molecule.
The concentration of V is an order of magnitude higher than that of VI, but

the evidence suggests that its destabilising effect on PVC is less than that of
36-39
VI. It has also been proposed that α ,β-unsaturated ketones, VII, are more po-
tent sensitisers for thermal degradation than
the α-chloroallyl groups in V and VI. Although the metal carboxylates do dis-
place chlorine from PVC during processing (reaction 7), it is not now clear how
important this reaction is in stabilising PVC in practice.
G. Scott 15
Allylic methylene is much more strongly implicated under the oxidative

conditions experienced by PVC in a screw extruder (see 3 above) and since the
unsaturated ketone, VII, is formed from the allylic hydroperoxide, VIII, it seems
probable that under processing conditions the latter rather than the former is
the primary cause of instability by reaction 6. A similar chlorine displacement
40
mechanism has been proposed for a variety of synergistic compounds (IX-XI).
6 ANTAGONISM AND SYNERGISM

It will be clear from the above that recycled polymers are normally unsuitable
for out-door use unless considerable care is taken in devising the correct formu-
lation for such possible applications as piping, bumpers, etc. The provenance of
the waste plastic is critically important since many synergistic heat stabilising
systems accelerate out-door weathering. Some sulphur antioxidants (e.g.
thiodipropionate esters) antagonise with the common hindered phenols under
41
these conditions and are generally antagonistic toward the hindered amine
41
class of light stabiliser. This makes the design of stabiliser formulations for re-
cycled engineering plastic extremely difficult unless the provenance of the
waste is known with some certainty. This is often made more difficult by the fact
that polymer manufacturers and fabricators are generally reluctant to disclose
the composition of their stabiliser “packages”. Blending of two identical poly-
mers containing different stabilising systems could lead to a recyclate with re-
duced stability due to antagonism between the components of the residual
stabilisers. Some stabilisers show antagonism in thermal stabilisation but this
is less common than during weathering.
In general, an effective light screening pigment such as carbon black or ti-

tanium dioxide will provide UV protection to the bulk of a thick-walled artifact
but this stratagem is not effective in thin films unless coupled with an effective
42
synergist. The dialkyl tin processing stabilisers I and II differ sharply in their
light stabilising ability; the maleate (I) being more effective than the
thioglycollate (II). However, when combined with a UV absorber, II become a
43
much more effective synergist than I. Polymer-bound sulphur synergists be-
42,43
have similarly.
The most satisfactory strategy for successful recycling of engineering poly-
mers to durable end products is to use only one source of plastic waste whose
composition and history are known. This then permits the design of a stabiliser
system which will synergise effectively with what remains of the previous stabi-
liser package or, at worst, will not antagonise with it.
7 ANTIOXIDANT-MODIFICATION OF POLYMERS DURING PROCESSING
The physico-chemical behaviour of antioxidants and stabilisers in polymers is
equally as important to their effectiveness as is their chemical activity. At its
simplest; if the stabiliser is lost from the polymer due to volatilisation or leach-
ing then it will not protect the polymer under aggressive technological condi-
tions, however effective it might be in a closed system. It has been shown that at
high temperatures, particularly in a moving air stream, many antioxidants are
44,45
physically lost by volatilisation. Similarly, at much lower temperatures, all
antioxidants and stabilisers are extracted from polymers by solvents, lubricat-
45-48
ing oils, and greases. An example of this behaviour in PVC plastisols is the
degradation that occurs in leathercloth, particularly in sunlight, due to the de-
pletion by oil leaching, evaporation, etc., initially of stabilisers and antioxidants
and subsequently of plasticiser, leading to cracking and embrittlement. Similar
effects occur in impact modified PVC subjected to leaching by the weather.
Diffusion of additives through the polymer bulk and subsequent loss by vo-
latilisation can be reduced in the above examples by incorporating the stabiliser
function in a large molecule which is soluble in the substrate. The shape and size
31
of an additive affects its rate of diffusion and rate of volatilisation and in the
case of the large compact molecules, both of these are minimised compared with
lower molecular weight compounds (e.g. IV, BHT, 1076, etc.).
Some attention has been directed to making antioxidants and stabilisers
46-49
part of the polymer molecule. Two main processes are used; copolymerisa-
G. Scott 17
tion during manufacture and grafting to an existing polymer chain. The first is
expensive compared with the use of conventional additives and can only be justi-
fied when antioxidant permanence cannot be achieved any other way. The sec-
ond is economically more attractive but generally suffers from the disadvantage
that conventional grafting normally gives relatively low yields of adduct due to
49
competition from homopolymerisation and homopolymerised vinyl antioxi-
50
dants are normally incompatible with the main polymer phase.
However, it has been found possible to use the relatively low levels of
monoeneic unsaturation formed in PVC during processing (Scheme 1) to give a
very high level of chemical binding of thiol antioxidants (e.g. XII, XIII) by the ad-
48
dition reaction shown in Scheme 4. Table 2 illustrates the effective synergism
between these polymer-reactive stabilisers and conventional processing stabi-
lisers, DBTM and DOTG in an accelerated weathering test when added during a
42,51
normal processing operation. Both thiols were found to be completely bound
51
to the polymer within the first few minutes of processing.
Table 2: Sulphur-containing antioxidants as photoantioxidants

in PVC. Total concentration 5.8×10-3 mol/100 g.48
Stabilizer Embrittlement time, h
DBTM (I) 1000
DOTG (II) 460
DBTM + BHBM (XII) 1400
DOTG + BHBM 400
DBTM + EBHPT (XIII) 1850
DOTG + EBHPT 1000
Control (no antioxidant) 670
An alternative approach to stabiliser substantivity under aggressive con-

ditions involves reactive processing in the presence of a peroxide. Symmetrical
derivatives of maleic acid, e.g. XIV, are very reluctant to homopolymerise but do
52
readily graft to PVC (see Scheme 5). Under the appropriate conditions, this
can be done without any degradation of the PVC.
Scheme 4: Antioxidant adduct formation in PVC during processing
Scheme 5: Maleic ester adduct formation in PVC
8 CONCLUSIONS
• Due to the high chlorine content of PVC some of the techniques at present
being considered in the “Systems” waste disposal options are unfavour-
G. Scott 19
able. In particular, degradability and composting are not suitable because

of the unknown hazards associated with the oxidative degradation of PVC
in the environment. Incineration and pyrolysis are similarly disfavoured
because of the large amounts of hydrogen chloride and other toxic products
that might be produced.
• Of the two acceptable disposal technologies, materials recycling and land-
fill, the former is preferred when the provenance of PVC waste is known.
Thus, clean PVC whose composition (particularly with respect to stabilis-
ers) and previous history is known is suitable for “closed loop” recycling in
non-foodstuffs applications.
• The problem of antioxidant and stabiliser loss from consumer products is
exacerbated by the recycling process but can be mitigated by designing the
stabiliser system to be an integral part of the polymer structure.
• The discipline of closed-loop recycling will require a new approach to the
design of polymer stabilisers in order to maximise synergism and minimise
antagonism between individual components of the stabiliser package
both-during initial processing and use and subsequent recycling. De-
signing PVC formulations for recycling rather than for a single use after
which the artifact is discarded involves a new approach to stabiliser design
with careful monitoring of the formulation at each stage whenever recov-
ered material is blended with virgin polymer.
ACKNOWLEDGMENT
I am grateful to colleagues and former students who have contributed to
the research that has made possible the conclusions reached in this review and I
thank in particular Dr. B. B. Cooray, Dr. X. Xing, and Dr. J. Li for the opportu-
nity to refer to previously unpublished results.
REFERENCES
1. G. Scott in Degradable Materials, Eds. S. A. Barenberg, J. L. Brash, R. Narayan,
and A. E. Redpath, CRC Press, p. 143, 1989.
2. G. Scott in Recycle’93, Proceedings of 6th Annual Forum on Recycling, Davos,
March 22-26, 1993, Maack Business Services.
3. H. Harata in Recycle’93, Davos, March 22-26, Paper 15/4, 1993.
4. D. Gilead and G. Scott in Developments in Polymer Stabilisation-5, Ed. G. Scott,
App. Sci. Pub., Chapter 4, 1982; D. Gilead, Polym. Deg. and Stab., 29, 63 (1990);
D. Gilead in Degradable Polymers, Principles and Applications, Eds G. Scott

and D. Gilead, Chapman & Hall, in press.
5. G. Scott in Proc. Second Int. Conf. on Marine Debris, Eds R. S. Shomura and
M. L. Godfrey, Vol. 1, U.S. Dept of Agriculture, p. 827, 1990; G. Scott, Polym. Deg. and
Stab., 29, 135 (1990).
6. G. Hanna in Recycle’93, Davos, March 22-26, Paper 18/2, 1993.
7. G. Scott in Biodegradable Plastics and Polymers, Proc. 3rd Int. Conf.,
Ed. Y. Doi et al., Elsevier Science Pub., in press; G. Scott in Degradable Polymers,
Principles and Applications, Eds. G. Scott and D. Gilead, Chapman & Hall, in press.
8. C. Sadrmohaghegh, G. Scott, and E. Setudeh, Polym. Plast. Technol. Eng., 24, 149
(1985).
9. N. Grassie and G. Scott in Polymer Degradation and Stabilisation, Cambridge
University Press, p. 105, 1985.
10. W. Kaminsky in Recycle’93, Davos, March 22-26, Paper 7/4, 1993.
11. G. Scott in Atmospheric Oxidation and Antioxidants, Ed. G. Scott, Vol. II,
Chapter 3, Elsevier Science Pub., 1993; G. Scott in Atmospheric Oxidation and
Antioxidants, Ed. G. Scott, Volume II, Elsevier Sci. Pub., Chapter 8, 1993.
12. G. Scott, Polym. Plast. Technol. Eng., 11, 1 (1978).
13. G. Scott, Resource Recovery and Conservation, 1, 381 (1976).
14. G. Scott, Proc. Int. Conf. on Advances in the Stabilisation and Controlled
Degradation of Polymers, Luzerne, May 1990, p. 215.
15. G. Scott in Developments in Polymer Degradation-1, Ed. N. Grassie, p. 205, 1977.
16. G. Scott, M. Tahan, and J. Vyvoda, Europ. Polym. J., 14, 913 (1978).
17. G. Scott, M. Tahan, and J. Vyvoda, Europ. Polym. J., 15, 51 (1979).
18. B. B. Cooray and G. Scott in Developments in Polymer Stabilisation-2,
Ed. G. Scott, App. Sci. Pub., p. 89 et seq., 1980.
19. F. Chevassus and R. De Broutelles in Stabilisation of Polyvinyl Chloride, Arnold,
p. 223, 1963.
20. G. Scott in Biodegradable Plastics and Polymers, Eds. Y. Doi and K. Fukuda,
Elsevier Science B.V., p. 307 et seq., 1965.
21. J. J. Gormley, Reg. Tech. Conf. SPI, March 24, 1970, p.63.
22. A. Ghaffar, A. Scott, and G. Scott, Europ. Polym. J., 11, 271 (1975).
23. A. Ghaffar, A. Scott, and G. Scott, Europ. Polym. J., 13, 83 (1977).
24. M. Ghaemy and G. Scott, Polym. Deg. and Stab., 3, 233 (1981).
25. M. Ghaemy and G. Scott, Polym. Deg. and Stab., 3, 253 (1981).
26. G. Scott and M. Tahan, Europ. Polym. J., 13, 981 (1977).
27. G. Scott in Developments in Polymer Stabilisation-1, Ed. G. Scott, App. Sci. Pub.,
p. 309, 1979.
30. A. Ghaffar, C. Sadrmohaghegh, and G. Scott, Europ. Polym. J., 17, 941 (1981).
31. G. Scott in Atmospheric Oxidation and Antioxidants, Ed. G. Scott, Elsevier Sci.
Pub., Vol. II, Chapter 5, 1993.
G. Scott 21
Pub., Vol II, p. 184 et seq., 1993.
33. K. S. Minsker, M. I. Abdullin, S. V. Kolesov, and G. E. Zaikov in Developments in
Polymer Stabilisation-6, Ed. G. Scott, App. Sci. Pub., Chapter 5, 1983.
34. S. Al-Malaika, K. B. Chakraborty, and G. Scott in Developments in Polymer
Stabilisation-6, Ed. G. Scott, App. Sci. Pub., p. 73, 1983.
35. B. B. Cooray and G. Scott, Polym. Deg. and Stab., 2, 35 (1980).
36. A. H. Frye and R. W. Horst, J. Polym. Sci., 40, 419 (1959).
37. K. S. Minsker, V. V. Listiskii, and G. E. Zaikov, Polym. Sci. USSR, 23, 535 (1981).
38. V. V. Lisitiskii, S. V. Kolesov, R. F. Grataullin, and K. S. Minsker, Z. Analit. Khimii,
33, 2202 (1978).
39. E. N. Zilberman, Perepletchikova, Y. N. Getmanenko, V. I. Zegelman, T. Molova, and
Y. A. Zvereva, Plast. Massy, 3, 9 (1975).
40. K. S. Minsker, A. A. Berlin, V. V. Lisitskii, and S. V. Kolesov, Vysokomol. Soed., 19, 32
(1975).
41. A. Guyot and A. Michel in Developments in Polymer Stabilisation-2, Ed. G. Scott, App.
Sci. Pub., p. 89, 1980.
Pub., Vol. II, Chapter 9, 1993.
43. B. B. Cooray and G. Scott, Europ. Polym. J., 17, 229 (1981); G. Scott in
Developments in Polymer Stabilisation-6, Ed. G. Scott, App. Sci. Pub., p. 67, 1983.
44. G. Scott, Pure & App. Chem., 30, 267 (1972).
45. N. C. Billingham in Atmospheric Oxidation and Antioxidants, Ed. G. Scott,
Second Edition, Elsevier Sci. Pub., Vol. II, Chapter 4, 1993.
Chapter 9, 1979.
Chapter 6, 1981.
48. G. Scott in Developments in Polymer Stabilisation-8, Ed. G. Scott, Elsevier App.
Sci., Chapter 5, 1987.
49. D. Munteanu in Developments in Polymer Stabilisation-8, Ed. G. Scott, Elsevier
App. Sci., Chapter 4, 1987.
50. B. W. Evans and G. Scott, Europ. Polym. J., 10, 453 (1974).
51. B. B. Cooray and G. Scott, Europ. Polym. J., 17, 385 (1981).
52. J. Li, G. Scott, and X. Xing, unpublished work.
D. Braun and K. Krämer 23
Investigations
on the Recycling Ability
of Used PVC
D. Braun and K. Krämer
Deutsches Kunststoff-Institut, Schloßgartenstraße 6, 64289 Darmstadt, Germany
INTRODUCTION
With a present world-wide annual production of approximately 19 million
tones, poly(vinyl chloride) (PVC) is the second largest volume thermoplastic
1
only to polyethylene as volume leader in the plastics industry. Its ability to be
compounded with many additives to a wide range of flexible and rigid forms con-
stitutes the major factor responsible for the versatility of PVC. Because of a low
cost and the processability by a wide variety of techniques (e.g. calendering, ex-
trusion, injection molding, and plastisol techniques) combined with good physi-
cal, chemical, and weathering properties, PVC has become a universal
2
polymer. There are many applications of rigid and plasticized PVC, e.g., pipes,
profiles, floor coverings, cable insulation, roofing sheets, packaging foils, bot-
tles, and medical products.
At the end of the service-time of these articles, large amounts of scrap
arises. Today, the question of the disposal of used plastics has gained increasing
importance in the public discussion because of the environmental problems re-
sulting from the rapid growth of the plastic waste during the last years.
Landfilling of municipal solid waste is becoming a burden as for example in the
United States about 80% of waste is dumped into landfills. A continuation at the
24 Investigations on recycling ability of used PVC
3
present rate could exhaust the U.S. landfill capacity in a decade. Also some Eu-
ropean countries are faced with a similar dilemma because the availability of
4
suitable sites is limited. Thus, landfilling as a disposal process is increasingly
seen as a last option.
The energy recovery by incineration is another way to dispose the munici-
5
pal solid wastes. But environmental argumentation, such as toxic emissions
from inadequate equipment or inappropriate incineration conditions, are build-
6
ing up a public resistance against these techniques. Especially, PVC incinera-
tion is connected with some technological problems due to the high chlorine
content of this polymer which yields large amounts of hydrogen chloride during
thermal decomposition, beside the possibility of formation of toxic dioxines and
furans. Therefore, plans to expand capacities of such installations meet with
growing difficulties.
The recycling activities can be separated to chemical and material recy-
cling. The chemical recycling is based on the idea of converting polymers back
into short-chain chemicals for re-use in polymerization or other petrochemical
processes. Four different process technologies are currently considered for
7
chemical recycling: cracking, gasification, hydrogenation and pyrolysis. In the
meantime, a few commercial-scale plants are working and some interesting
8-10
studies about investigations with these techniques are available. At the end,
the economic efficiency will be decisive for the application as any recycling pro-
cess in the future.
The material recycling is already practiced in plastics industry over many
years in regard to post-manufacturing waste. These experiences can be used to
develop new concepts for material recycling of post-consumer waste. The major
problem in the recycling of used plastics is connected to a great inhomogeneity of
11
such polymers present in the waste. A statistical study by the Information Sys-
tem on Plastics Recycling in Western Europe shows that about 7.4% of the 9 mil-
lion tones of municipal solid waste in Western Europe in 1990 are plastics
materials. Figure 1 illustrates the percentage of different polymer types in the
total plastics portion.
The incompatibility of these components is the reason of the difficult pro-
cessing and inferior mechanical properties of the resulting products from mixed,
chemically different polymers. Therefore, it is necessary to separate various
polymers to boost their value. Although, there are some practical problems,
12,13
some interesting developments for plastics waste separation were found.
The separation in a hydrocyclone, which works based on the principle of sorting

by a centrifugal force field, using density difference of the various polymers is
14
one possible solution. Sometimes, prior to separation, it becomes necessary to
clean the polymer waste to remove contamination like dirt, food, and paper.
Figure 1. Percentage of different polymer types in the total plastics portion of municipal solid
waste.
In the future, a clean PVC-stream of the municipal solid wastes and the
building sector wastes ready for material recycling can be expected. This evolu-
tion will be accelerated by new ordinances made by the government of some
countries. In Germany the first item of legislation to be introduced was the “Act
15
on the Avoidance of Packaging Waste”. The aim of this legislation is to reduce
the large amounts of packaging through avoidance and material recycling. An-
other fact is that the manufacturers and traders are made responsible for their
used packaging to relieve the local authorities of the burden to dispose the
waste.
Generally, one can say that material recycling is a necessary way to reduce
the municipal solid waste problem. This paper shows a concept of integrated
steps required to make recycling happen. The special problems of PVC recycling
will be presented and illustrated by a number of practical examples.
CHARACTERIZATION OF PVC WASTE
Many different grades and types of PVC are available allowing applica-
tions as diverse as flexible sheets, pressure pipes, transparent bottles, and med-
ical products to be produced. For these articles, a lot of different additives and
stabilizer systems are used to get suitable properties for the respective applica-
16
tions. Moreover, during a high temperature processing and throughout a ser-
17
vice live of the product the polymer might be subjected to degradation.
Therefore, a characterization of the PVC waste is necessary to obtain informa-
tion on properties such as the residual stability, molecular weight, and content
of additives of the individual PVC species.
MEASUREMENT OF RESIDUAL STABILITY
The main disadvantage of PVC is the rather limited thermal stability

which requires addition of heat stabilizers to prevent dehydrochlorination and
discoloration. With respect to a great practical importance of the polymer, the
thermal and photochemical degradation of PVC has been studied for a long time
18-21
and there is a large number of published surveys. The elimination of hydro-
o
gen chloride, at relatively low temperatures (about 100 C) or under the influ-
ence of light is the fundamental aspect of PVC decomposition. In the first stage,
this reaction leads to the formation of one double bond followed by a so-called
rapid zipper-like splitting off of further HCl molecules to give polyene sequences
(Figure 2). These sequences, with a mean length of 6 - 14 conjugated double
bonds, cause the polymer to turn yellow, brown, and eventually black.
Figure 2. Schematic formula of the dehydrochlorination of PVC.
The thermal stability of PVC is considerably lower than that of its low-mo-
lecular weight model analogues. Initial sites, such as allylic chlorines adjacent
to internal double bonds, tertiary chlorines at branched carbons, head-to-head
units, and oxygen-containing structures are believed to be responsible for the in-
stability. The mechanisms, which occur during degradation are not yet fully un-
derstood. There are radical or ionic mechanisms suggested, and it seems that
the type of reaction depends on the conditions (temperature, presence of oxygen,
etc.) during the decomposition.
he main function of heat stabilizers is to prevent degradation during pro-

cessing. They have in common the ability to react with HCl when it is liberated
from the polymer. Another task is to replace labile chlorines, which may initiate
the dehydrochlorination of more stable groups, and thus to enhance the heat
stability. A number of organometallic compounds and inorganic salts are espe-
cially effective.
A part of the stabilizer will be consumed during processing and sometimes
during the application period. Therefore the efficiency of the stabilizer system is
21
remarkably reduced. That makes it necessary to get information about the re-
sidual stability of PVC articles before they can be recycled.
For this purpose, the determination of the hydrogen chloride elimination
seems to be the best way. The study of the early stages of the reaction requires a
combination of good reproducibility, high accuracy and a low detection limit.
The following picture (Figure 3) shows an apparatus which is very suitable and
23
often used for such studies.
Figure 3. PVC degradation measuring apparatus. a. rotameter, b. degradation vessel with PVC
sample, c. thermostat (180oC), d. conductivity cell (25oC), e. conductivity-meter, f. computer.
The PVC-sample (≈ 0.1 g) is introduced into the degradation vessel and

o
then the measurement is carried out under isothermal conditions (e.g., 180 C).
A stream of warmed up carrier gas (nitrogen) transports the evolved hydrogen
chloride into the conductivity cell filled with distilled water. The HCl determi-
nation is performed by continuous conductometric measurements. As a result

the conversion-time-curve is obtained as illustrated in Figure 4.
Figure 4. Schematic degradation curve of stabilized PVC.
The degradation curve of stabilized PVC shows an induction period where

no HCl is evolved. During this period, the heat stabilizer is consumed and after-
wards the dehydrochlorination begins. The time of induction, ti, is an important
information required to estimate the remaining stability of a PVC specimen and
decide whether an additional stabilization is necessary for the material recy-
cling. Also, the rate of HCl split-off, after stabilizer’s consumption, can be calcu-
lated. In some cases, it can be sufficient to use a simple Congo Red test, e.g.,
according to DIN 53418, instead of the apparatus for measuring the hydrogen
chloride elimination.
ANALYSIS OF THE HEAT STABILIZER AND OTHER ADDITIVES
As mentioned above, a limited thermal stability of PVC requires the addi-

tion of heat stabilizers in almost all fields of application. Besides, also other ad-
ditives (e.g., light stabilizers, fillers, lubricants) are used to modify the
properties of PVC or to improve its processability. At present, about 1/3 of all
24
used PVC is plasticized by various types of modifiers. Therefore, it is helpful to
get some detailed information about the composition of a special PVC scrap be-
fore re-use.
This can be done by some analytic methods based on the experiences of the
Deutsches Kunststoff-Institut (DKI) with the material recycling of various used
PVC products. As an example, for plasticized PVC, the analysis of PVC roofing
25
sheets is described and shown in Figure 5.
Figure 5. Analysis of PVC roofing sheets.
The first step consists a Soxhlet-extraction of the powdered PVC sample

with diethyl ether to separate plasticizer. After evaporation of the solvent, the
amount of plasticizer can be determined. The rest of the material is then dis-
solved in tetrahydrofurane (THF), and, after filtration, the fibrous materials are
obtained. The other components, insoluble in THF, are separated in a centri-
fuge. The remaining residue can be divided to fillers and crosslinked PVC by
burning to ashes. By dropping the THF solution in a surplus of methanol the
PVC is precipitated. The single components are determined gravimetrically and
identified by chemical and spectroscopic methods. Usually the quantitative
analysis of the main parts of a PVC sample (plasticizer, filler, PVC itself) will
give enough information about the material. For the qualitative analysis the
IR-spectroscopy is particularly suited because the main additives, including co-
26
polymers and impact modifiers, have typical IR-bands. Also other spectro-
scopic methods can be used for identification but the expenditures of sample
preparation and equipment are higher. A complete qualitative and quantitative
27
analysis of all ingredients of a PVC compound was previously discussed.
Finally, the determination of the heat stabilizer, as an important point in
the analysis of PVC waste, is particularly considered. The selection of a stabi-
lizer system for PVC depends on many factors including application, tradition of
28
the market, and local legislation. Lead stabilizers are the most widely used
PVC heat stabilizers because they provide cost-effective stabilization systems
and easy processing. They maintain volume resistivity in plasticized PVC cable
insulation and are the principal stabilizers for many general-purpose applica-
tions. Several metal carboxylate soaps are used in combination as PVC stabiliz-
ers, e.g., barium-cadmium, barium-zinc, calcium-zinc. Since many years,
barium-cadmium systems were used in Europe in white window frames with
good weathering properties. But the utilization of cadmium in stabilizers or pig-
ments recently comes under increasing scrutiny. At present, all manufacturers
29,30
are looking for alternative systems such as calcium-zinc stabilizers. They
are applied for food packaging, water bottles, and medical products. The number
of applications is likely to increase with the availability of less toxic additives.
The organotin compounds form another large group of stabilizer systems where
mono- and dialkyltins are the most widely used. Their properties depend on the
nature of the alkyl and acid groups present. The toxicity of the dialkyltin de-
creases rapidly with the chain length of the alkyl group, so that octyl tin com-
pounds are accepted for food contact applications. Also, some sulfur-containing
organotin-stabilizers are used because of their excellent heat stability and clar-
ity. The relatively high costs is the main drawback of tin stabilizers.
A simple possibility to obtain a detailed information about the stabilizer
system in PVC waste can be seen in the classic analysis methods which are com-
mon practice in inorganic chemistry of the separation and determination of cat-
ions. The only difficulty is to find an easily practical way to get the metallic
cations into water phase. For this purpose, the PVC sample is dissolved in
cyclohexanone and the received solution used for a liquid/liquid-extraction with
nitric acid containing water. After phase separation, the different cations are
found in water solution.
The determination of the metals can also be accomplished by thin layer

31
chromatography, using an organic solution of PVC in THF. Sometimes, a pre-
cipitation of the polymer might be necessary, and the remaining methanol/THF
solution is used for the identification. Besides, some spectroscopic methods are
32
described for stabilizer analysis. The infrared spectroscopy seems to be the
preferred method for this purpose because of its easy feasibility combined with a
33
high detection rate.
DETERMINATION OF MOLECULAR WEIGHT
For different kinds of processing and various applications, industry offers

34
PVC types with K-values between 55 and 80. The K-value is a traditional unit
of measurement used by manufacturers to describe the molecular weight of PVC
materials. This information is necessary to decide which processing technique
can be used for recycling. Also, under the influence of heat, light, and oxygen,
PVC chains can be degraded or even crosslinked which results in changes in the
22
molecular weight and molecular weight distribution. There is a correlation be-
tween molecular weight, processability, and mechanical properties of PVC, and
it is thus important to investigate changes in molecular weight during process-
ing or use.
The simplest method for molecular weight measurement includes the de-
termination of the viscosity of a PVC solution. The PVC is usually dissolved in
o
cyclohexanone and measured at 25 C, e.g., according to DIN 53726. For practi-
cal purposes, the obtained K-value gives sufficient information in most cases.
Using the Mark-Houwink-equation, the molecular weight can also be calculated
35
from the results of the viscosity measurements.
The gel permeation chromatography, GPC, is by far the most popular
21
method of molecular weight measurement. It not only gives information on the
molecular weight but also on the molecular weight distribution. Normally, THF
is used as solvent and the columns are calibrated with polystyrene or PVC stan-
dards. In some cases, the results might be misleading, either if PVC is not dis-
solved properly or if the sample contains polymer that is partially insoluble due
to former treatment, e.g., crosslinking under the influence of heat or light.
POSSIBILITIES OF ADDITIONAL STABILIZATION OF USED PVC

The recycling of PVC waste offers a lot of problems due to the limited ther-
mal stability which requires, in most cases, the addition of stabilizers. PVC arti-
cles are susceptible to degradation at almost all stages of their lifetime:
production, storage, processing, transportation, and end-use. Several possible
influences, which can be hostile to polymers, are heat, light, oxygen, and me-
chanical stress. Therefore the stabilizer system is partially consumed during
the service life of PVC product. An example is shown in Figure 6 where the con-
version-time-curve of a 15 - 20 years old window frame scrap and a post-manu-
facturing window frame waste are compared.
Figure 6. Degradation curve of a post-manufacturing window frame waste and an old window
frame scrap (180oC, nitrogen); post-manufacturing waste: —-; window frame scrap: ---.
The induction time of the post-manufacturing waste is much longer than

the induction period of the old PVC scrap. Also, some other studies about the
changes in properties of rigid PVC during weathering show a reduction of the
36
stabilizer’s efficiency.
ADDITION OF HEAT STABILIZERS
As mentioned before the measurement of the residual stability of a PVC ar-

ticle may lead to the result indicating that an additional stabilization is neces-
sary for efficient recycling. The use of a new heat stabilizers for this purpose is
connected with some difficulties. First, it is necessary to determine the stabi-
lizer system present in the PVC waste because some stabilizers are not compati-
ble with others. For instance, a sulfur-containing tin stabilizer reacts with a
lead stabilizer during processing and the resulting material would have dark
spots from lead sulfide formed, besides the reduced efficiency of the system.
Moreover, the use of heavy metals in stabilizers might be forbidden in a few
years by the governments of some European countries because of their toxicity.
If this happens, every processing company will have the problem how to stabi-
lize old PVC products if material recycling is desired.
IMPROVEMENT OF THERMAL STABILITY BY FILLERS
The use of a filler as a co-stabilizer is an alternative possibility to recycle

PVC waste without addition of further stabilizers. For this purpose, calcium car-
bonate is suitable, because it is able to react with hydrogen chloride. Further-
more, chalk has good properties (e.g., wide variety of calcium carbonate
materials available, low price, no abrasion of processing equipment, reduced
plate-out, increasing mechanical properties, and homogeneous distribution by
16
coating) making it popular as PVC filler. In Figure 7, the degradation-curves of
two stabilized PVC samples (one with 10 phr (parts per hundred parts resin) cal-
cium carbonate and one without filler) are illustrated.
The addition of chalk increases thermal stability of the filled PVC sample,
indicated by a longer induction period of dehydrochlorination. Also, the rate of
HCl-elimination is lower in the presence of calcium carbonate. Other investiga-
tions with unstabilized PVC samples have shown that the filler acts as a trap for
the split-off hydrogen chloride, but it has no influence on the decomposition of
PVC. This is confirmed by the UV-spectra of PVC which was heat-treated in the
22
absence and in the presence of chalk. Under both conditions, the same unsatu-
rated sequences in PVC are formed, as shown in Figure 8 (curve a and b). On the
contrary, in the presence of stabilizers (curve c) practically no polyene sequences
are formed during the induction period.
Consequently, one can conclude that chalk acts as a co-stabilizer for PVC
increasing its thermal stability, but only in connection with a normal heat stabi-
lizer or the remaining active heat stabilizer in PVC waste. The lower HCl-elimi-
nation rate can be explained by the fact that hydrogen chloride, which reacts
with the filler, has no more the well-known catalytic effect on the further PVC
37
degradation.
Figure 7. Dehydrochlorination of PVC at 180oC under nitrogen. PVC filled with 10 phr chalk —-,
PVC without filler ---.
RESULTS OF PROCESSING EXPERIMENTS WITH FILLED PVC

Some studies with filled and unfilled rigid PVC were made in our labora-
tory to describe the effect of chalk as additional stabilizer. A suspension PVC
(K-value 70) with an organotin stabilizer (2 phr) and lubricants (1.8 phr) was
used as the experimental material. A part of this compound was filled with a
stearic acid coated calcium carbonate (10 phr) as an additional component. The
PVC powder and the additives were mixed at a high speed in an intensive mixer.
The received dry blends were pelletized by extrusion to get a better dispersion of
the additives in the polymer material. Finally, the granulates were processed by
injection molding to test specimens for measurements of mechanical properties.
Tensile stress-strain data at constant deformation rate and constant tem-
perature are undoubtedly the most valuable mechanical data for the character-
38
ization of rigid PVC. They are widely used not only for material selection but
Figure 8. UV-VIS-spectra of PVC in THF (2 g/l. before (d) and after (a,b,c) thermal degradation (30
min at 180oC under nitrogen).
also to determine the strength of the resulting products and to measure the re-
tention of mechanical strength on outdoor or accelerated exposure.
Stress-strain measurements are generally made in tension, e.g., according to
DIN 53455. The standard test piece is stretched at a uniform rate until it breaks.
From the x-y plot, the tensile strength, the elongation at break, and the modulus
of elasticity are obtained. The toughness is a further important mechanical
38
property of a polymer material which can be determined by an impact test. The
testers are pendulum instruments that break the specimen with a hammer. For
this purpose the Charpy-method, according to DIN 53735, with notched stan-
dard test pieces can be used.
Table 1: Results of the mechanical measurements of the filled and unfilled PVC,
according to DIN 53455 and DIN 53735
Modulus Tensile Elongation Notched

of elasticity strength at break impact strength
2 2 2
N/mm N/mm % kJ/m
PVC without chalk 3190 66.1 10.5 4.3
PVC + 10 phr chalk 3360 62.4 11.0 5.0
The results of the mechanical measurements are given in Table 1. The

stress-strain-diagrams of the PVC samples show a ductile behavior with a
yield-point. The addition of chalk causes a higher modulus of elasticity, whereas
the tensile strength is slightly diminished. The elongation at break is nearly
equal, and also the notched impact strengths of the filled and unfilled PVC are
on the same level. We can infer from this data that the use of calcium carbonate
as filler brings no discredit upon the mechanical properties of rigid PVC if the
filler is homogeneously distributed in the polymer material during processing.
Figure 9 illustrates the degradation-curve of the two PVC samples. The in-
duction period of the filled PVC (≈ 300 min) is significantly longer than the time
of induction of the unfilled PVC (≈ 160 min). Also of interest is the fact, that, in
the presence of chalk, the rate of HCl-elimination after consumption of the stabi-
lizer system is lower. A comparison of Figure 9 with Figure 7 shows a corre-
sponding shape of the degradation-curves of the filled and unfilled PVC
samples. The additional stabilization of calcium carbonate in PVC compounds
could be proved also under realistic processing conditions.
Some more investigations, such as variation of the concentration of filler
and change of the particle size, were made to get further details about the influ-
ence of chalk in PVC. It could be shown that the stabilizing effect increases with
the content of calcium carbonate in the mixture. But there is a maximum at
about 30 phr filler because of the increase in shear viscosity in the used process-
ing machines which leads to an increasing thermomechanical treatment of the
material. Another limiting factor is the change of the mechanical behavior from
ductile to brittle with rising amount of chalk connected with lower values of
elongation at break and impact strength.
Figure 9. Conversion-time curve of the two processed PVC samples; Stabilized PVC with 10 phr
chalk —-; Stabilized PVC without filler ---.
PRACTICAL EXAMPLES OF PVC RECYCLING

To confirm the applicability of the suggested recycling concept with real
PVC waste, three different used rigid PVC products were investigated as de-
scribed previously. The obtained results are shown in Table 2.
The three samples were processed according to method used for the virgin
PVC as described above. First, the big pieces of PVC waste had to be handled in a
mill to get a powdered material. Then, half of the single regrinds were mixed
with 10 phr of a stearic acid coated calcium carbonate. After extrusion and injec-
tion molding the test specimens were received. Detailed information about the
processing conditions of the modification of PVC scrap with chalk, aiming at a
39
good distribution of the filler, are given elsewhere.
In Table 3, the results of the mechanical measurements of the three pro-
cessed PVC wastes without addition of filler are compared with the chalk modi-
fied samples. The influence of the added amount of chalk corresponds to the
effects discussed earlier, except for the notched impact strength. These values
are slightly higher for the modified specimens of the materials 1 and 3, whereas
the filled sample of material 2 has a lower value.
Table 2: Results of characterization of three different PVC wastes

Material 1 Material 2 Material 3
Article sheet foil window frame
Color grey colorless white
PVC (%) 87 93 87
Filler (%) 7 - 8
Other additives (%) 6 7 5
Stabilizer system Pb Ca/Zn Ba/Cd
K-value 61.5 60.2 69.3
Induction time (min)* 260 70 95
*180oC, nitrogen
Table 3: Mechanical measurements of produced specimens, according to DIN

53455 and DIN 53735. Upper value: processed PVC waste without chalk; Lower
value: processed PVC with 10 phr chalk
Modulus of elasticity 2850 2530 2500
(N/mm2) 2880 2730 2610
Tensile strength 57.6 50.3 50.2
(N/mm2) 53.3 48.0 46.5
Elongation at break 10.1 15.4 20.0
(%) 12.8 15.1 20.9
Notched impact strength 8.3 36.6 15.2
(kJ/m2) 12.1 25.9 22.9
The following Table 4 shows the induction time of the processed samples
and the original PVC waste.
Table 4: Determination of residual stability at 180oC under nitrogen (induction

time in min)
Original PVC waste 260 70 95
Processed PVC waste without addition of chalk 120 35 40
Processed PVC waste with addition of 10 phr chalk 170 95 65
Figure 10. Degradation curves of material 2; Original PVC waste —-; Processed PVC waste with-
out addition of chalk ....; Processed PVC waste with additional 10 phr of chalk - - -.
As expected the period of induction decreases due to the thermomechanical

treatment during processing, but a comparison of the unmodified and the modi-
fied PVC wastes indicates a considerable stabilizing effect of chalk. Also, the vi-
sual evaluation of the produced specimens indicates a better color quality of the
modified samples. Especially PVC scrap with a low residual stability (material 2
and material 3) needs an additional stabilization if intended for a re-use. This is
also impressively illustrated by the conversion-time-curve of material 2 (Figure
10). The chalk modified sample has a better heat stability than the original
product in spite of one more processing cycle.
A feasibility of PVC recycling without additional stabilization is seen in

mixing of PVC waste with a new material. Several studies are available about
40-43
the use of PVC scrap in the manufacturing process of various PVC articles.
They describe mainly the influence of regrind on the properties of virgin PVC.
The investigations made with the re-use of roofing sheets show that new PVC
roofing sheets can contain up to 10 - 20% of recycled material without any ad-
25
verse effect on the product quality.
New processing techniques offer another way to recycle PVC scrap. One ex-
44
ample is the production of window frames by coextrusion. The regenerated
PVC is used in core whereas virgin resin is used as the skin. In this process only
1/3 of new material is necessary to obtain the same properties as a window frame
from 100% virgin PVC. The coextrusion process is also feasible for the produc-
tion of pipes where the inside and the outside layers are made out of a new resin
45
and the old material is used for the thick middle layer.
CONCLUSIONS
• The studies on the recycling-ability of used PVC show the importance of a
careful characterization, especially if mixed PVC from the municipal solid
waste separation should be recycled. The analysis of the composition and a
detailed knowledge of the heat history and the molecular weight of a PVC
scrap are necessary before reprocessing can be applied.
• The suggested additional stabilization with calcium carbonate is an effec-
tive method for the recycling of PVC waste with a low remaining
thermostability. The addition of chalk, up to a content of 10 phr, does not
significantly change the mechanical properties, whereas it remarkably in-
creases the heat stability.
• A number of schemes have been initiated to collect plastics and to reprocess
them to useful articles. The most advantageous situation occurs when a
source of a single material type can be identified, for example, particular
types of packaging, bottles, or window profiles. Such materials can often be
simply reformulated and converted into high quality products.
• Nevertheless, new material recycling concepts for PVC mixtures have to be
established because in future the environmental preferences will play a
larger role in a material’s selection. A practical re-use of PVC requires a
continuous stream of suitable scrap and the further development of tech-
nologies to reach the specifications for the intended applications. Another
important point is that markets for the secondary products must exist to
make the material recycling a successful economical enterprise.
REFERENCES
1. I. P. Coates, Eng. Plast., 5, 362 (1992).
2. K. H. Felger, Ed., Kunststoff-Handbuch, Vol. 2/1, Polyvinylchlorid,
Hanser-Verlag, München, Wien, 1986.
3. J. W. Summers, B. K. Mikofalvy, G. V. Wooton, and W. A. Sell, J. Vinyl Technol.,
12, 154 (1990).
4. V. Matthews, Plast. Rubber Compos. Process. Appl., 19, 197 (1993).
5. L. R. Brecker, Polym. Prepr., 32(2), 125 (1991).
6. W. Freiesleben, Role and Importance of Material Recycling in Plastics Waste
Management during the Next Decade in Western Europe, G. Menges,
W. Michaeli, and M. Bittner, Eds., Recycling von Kunststoffen, Hanser Verlag,
München, Wien,1992.
7. P. Mapleston, Mod. Plast. Ind., 23(8), 32 (1993).
8. R. Holighaus and K. Niemann, Kunststoffberater, 38(4) 51 (1993).
9. M. W. Meszaros, Eng. Plast., 5, 335 (1992).
10. Anonymous, Neue Verpack., 45(10), 58 (1992).
11. F. P. La Mantia, Polym. Degrad. Stabil., 37, 145 (1992).
12. Anonymous, Gummi, Fasern, Kunstst., 44, 38 (1991).
13. R. T. Gottesman, Makromol. Chem., Macromol. Symp., 57, 133 (1992).
14. A. Schimpf, Kunststoffe, 82, 908 (1992).
15. M. Bittner and W. Michaeli, Annu. Tech. Conf. Soc. Plast. Eng., 252 (1992).
16. R. Gächter, H. Müller, Ed., Taschenbuch der Kunststoff-Additive, 3rd Ed., Hanser
Verlag, München, Wien, 1989.
17. S. H. Hamid, M. B. Amin, A. G. Maadhah, and A. M. Al-Jarallah, Annu. Tech. Conf.
Soc. Plast. Eng., 215 (1992).
18. D. Braun, Thermal Degradation of Poly(vinyl chloride), N. Grassie, Ed.,
Developments in Polymer Degradation-3, Applied Science Publishers Ltd., London,
1981.
19. D. Braun, Recent Progress in the Thermal and Photochemical Degradation of
Poly(vinyl chloride), G. Geuskens, Ed., Degradation and Stabilization of
Polymers, Applied Science Publishers Ltd., London, 1975.
20. T. Kelen, Polymer Degradation, Van Nostrand Reinhold Company, New York, 1983.
21. J. Wypych, Polyvinyl Chloride Degradation, A. D. Jenkins, Ed., Polymer Science
Library 3, Elsevier, Amsterdam, 1985.
22. D. Braun, Makromol. Chem., Macromol. Symp., 57, 265 (1992).
23. D. Braun and M. Thallmaier, Kunststoffe, 56, 80 (1966).
24. H. H. Meyer, Kunststoffe, 83, 744 (1993).
25. J. Deinert, Diploma Thesis, TH, Darmstadt, 1991
26. W. Kimmer and R. Schmolke, Plaste Kautsch., 32, 145 (1985).
27. D. Braun and E. Bezdadea, Angew. Makromol. Chem., 113, 77 (1983).
28. H. F. Mark, N. M. Bikales, C. G. Overberger, and G. Menges, Eds., Encyclopedia of

Polymer Science and Engineering, Supplement Volume, 2nd Ed., John Wiley &
Sons Inc., New York, 1989.
29. R. Beck, Plastverarbeiter, 39(10), 128 (1988).
30. E. Görlich, Kunststoffe, 80, 432 (1990).
31. E. Richter, Diploma Thesis, TH, Darmstadt, 1994.
32. D. Hummel and F. Scholl, Atlas der Polymer- und Kunststoffanalyse, Vol. 3,
Hanser Verlag, München, Wien, Verlag Chemie, Weinheim, 1981.
33. R. Schmolke, W. Kimmer, and W. Sauer, Acta Polym., 30, 432 (1979).
34. A. Franck and K. Biederbick, Kunststoffkompendium, 3rd. Ed., Vogel Verlag,
Würzburg, 1990.
35. M. Kurata and Y. Tsunashima, Viscosity-Molecular Weight Relationships,
J. Brandrup and E. H. Immergut, Eds., Polymer Handbook, 3rd Ed., John Wiley &
Sons Inc., New York, 1989.
36. I. Jakubowicz and K. Müller, Polym. Degrad. Stabil., 36, 111 (1992).
37. K. Krämer, Diploma Thesis, TH, Darmstadt, 1991.
38. I. L. Gomez, Testing rigid PVC Products, L. I. Nass, Ed., Encyclopedia of PVC,
Vol. 3, Marcel Dekker Inc., New York, 1977.
39. M. Speckhardt, Diploma Thesis, TH, Darmstadt, 1992.
40. J. W. Summers, B. K. Mikofalvy, H. K. Boo, J. M. Krogstie, J. C. Rodriguez, and
W. A. Sell, J. Vinyl Technol., 14, 166 (1992).
41. T. Fazey, J. Vinyl Technol., 14, 83 (1992).
42. Anonymous, Neue Verpack., 44(2), 77 (1991).
43. E. B. Rabinovitch and P. C. Booth, J. Vinyl Technol., 12, 43 (1990).
44. N. Berndtsen and V. Hofmann, Kunststoffe, 81, 907 (1991).
45. R. Saffert, Swiss Plast., 12(1/2), 71 (1990).
H. Uhlen 43
Recycling of Complete PVC Windows
Herbert Uhlen
VEKA AG, Postfach 1262, Sendenhorst, Germany
INTRODUCTION
Three basic materials are used for window frames: wood, aluminum, and
PVC. For approximately 30 years window frames of PVC have been used in Ger-
many. Even if the products of the first generation were not absolutely perfect,
PVC windows have still achieved a substantial percentage in the marketplace
because of their advantages: they are long-lasting, have low maintenance re-
quirements, and use only small amounts of material. In Germany, they have a
market share of 45%; this means 10,000,000 units are installed per year. It is ex-
pected that the share of PVC-windows will remain the same in the next few
years.
The durability and low maintenance is connected with the fact that PVC
does not rot. At the same time, even long-lasting building-products will fall into
disuse at some time, thus the question of waste management of old windows will
arise. Landfill disposal demands space and is not desirable. The incineration of
frames is also not desirable. The waste management of old plastic windows is
currently not sorted out. Some disposal sites accept the waste without problems
and due to the durability of the material the number of old windows is currently
still low.
44 Recycling of complete PVC window
As windows are quite big building parts, a separate collecting system is

worthwhile. Concerning the percentage of PVC in window frames, the recycling
to retain plastic material in the material chain is cost effective.
DISMANTLING SYSTEM
Table 1: Typical formulation of PVC for windows
Component Amount, parts

S-PVC 100
Impact modifier (acrylic resin) 7
Processing aid 1.5
Stabilizer (Pb-compound) 4
Stabilizer (Ba/Ca compound) 0.3
Lubricant 1.5
Extender (chalk) 6
Pigment 4
Complete windows are composed as follows: 38% PVC, 34% glass, 24%
metal, 4% other material. A window contains on average approx. 18 kg of PVC.
The plastic fraction of the window frame is a mixture of plastic and various
additives. A typical formulation is reported in Table 1.
For a long time, it has been common practice to grind the profile segments
remaining after cutting profile to size. Regrind obtained from this source is typi-
cally used in manufacturing of new profiles.
There have been attempts to refurbish old windows. Panes of glass were re-
moved, metal parts were taken off, gaskets were taken out, the plastic frames
were cut open and reinforcements were removed out of the profile pieces. This
method is labor intensive and, therefore, expensive.
At VEKA we have come to the conclusion that only a fully automatic dis-
mantling of the complete old windows can be made economical, considering that
the number of returned old windows will grow in the near future.
H. Uhlen 45
Figure 1. Automatic dismantling system.
We have taken into consideration the reprocessing of the numerous plas-

tic-covered wooden windows. These windows were installed in East-Germany
and have to be renewed in the next few years. Therefore, we have included the
separation of wood as a processing step in our plans.
Additionally, we are also carrying out a color separation. Most of the plastic
windows are pigmented white. After regrinding PVC into clean granules, we
separate the white grains from the non-white grains.
The automatic dismantling system is reported in Figure 1. The coarse
crushing takes place in a shredder of the type which is used for automobiles.
The coarse mixture of PVC, metal, glass, rubber, and possibly wood is pass-
ing through a rotating magnetic drum. The iron-containing metal parts are ex-
tracted and reach via a conveying belt the containers for sale.
The iron-free mixture goes afterwards into a sieve drum where material is
separated according to particle size into 4 fractions, as follows:
• particle size of <4 mm: mainly consisting of fine particles of glass (the
amount is approx. 2.5 %) cannot be currently economically utilized
• particle size of 4-15 mm is stored intermediately in silos and will undergo
subsequent processing
• particle size of 15-45 mm is conveyed to the non-iron separator
• particles above 45 mm are returned to the shredder to be crushed again.
In an eddy current separator non-iron metals are removed and stored.
Figure 2. Separation.
A similar eddy current separator operates in the line to process the mate-
rial of 4-15 mm particle size.
In this line, after metals were separated, we have a mixture of PVC, glass,
rubber, and possibly splinters of wood.
The rubber separator removes most of the rubber. The particles of the mix-
ture fall on a rotating cylinder which is covered with rubber. Particles made out
of rubber stop motion because of friction and fall down vertically. Particles made
of PVC, glass or wood bounce off and fall down further away.
After the rubber has been removed the mixture is conveyed back to the
shredder to be crushed again so that particle sizes of 4-15 mm is obtained.
The resulting mixture having particle size of 4-15 mm and consisting of
PVC, glass, residues of mortar, possibly splinters of wood reaches another sieve
drum (Figure 2). In this sieve drum a separation is carried out into two ranges of
particle size 4-8 mm and 8-15 mm.
For each fraction there is a bunker for intermediate storing. From each
bunker two parallel processing lines begin (all together four processing lines). In
each of these four lines, separation of wood and glass is carried out.
The separation of wood takes place on vibrating tables with sloping sur-
faces. Air is blown through holes in the tables. The lightweight wood particles
H. Uhlen 47
Figure 3. Sorting.
move downwards on the air stream whereas the heavy particles, PVC and glass,
are carried away upwards. The wood fraction is conveyed into containers.
The separation of glass particles takes place on similar vibrating tables.
The lightweight particles (PVC) move downwards on the air stream
whereas the heavy particles (this time: glass) are carried away upwards. The
glass fraction is conveyed into containers, too.
The PVC fraction free of wood and glass, is gathered together from all the
four processing lines (the two particle sizes range 4-8 mm and 8-15 mm are
mixed again).
Particles of metal which accidentally have not yet been sorted out are now
found by an electric coil for metal detecting (Figure 3). Each portion of PVC in
which a metal particle is found, is also collected. The material collected in such
way is put into the separation process again from time to time. Only PVC which
is absolutely free of metals undergoes the further process.
The PVC particles are crushed by means of a cutting mill to particle sizes of
2-8 mm. In the cutting mill the material is washed with water for the first time.
Another washing of the granules is carried out subsequently, see Figure 3.
A drying process with heated air is carried out in two steps. By means of
sieves, the PVC granules are then separated into 4 fractions having particle
sizes of <2 mm, 2-4 mm, 4-6 mm, and 6-8 mm. Small quantities of grains above 8
mm undergo the milling process again. For each fraction a final separation of
rubber is carried out. On vibrating tables, fine particles of rubber are sorted out.
Each fraction then undergoes a process of separation according to color
where white PVC grains are separated from non-white grains. A beam of light
touches every single grain. The reflected light is detected so that every grain is
recognized as being white or non-white. Non-white grains which pass the detec-
tor are deflected by air pulses while falling whereas white grains simply fall
down.
The dry PVC granules are stored in five big outdoor silos. Four of them con-
tain white PVC sorted according to particle size. The other one contains
non-white PVC.
The PVC granules are suitable for the production of new profiles. It is ad-
vantageous to have grains of the same size, therefore, we produce four ranges of
particle size as for the white granules. As for the non-white granules separation
according to particle size is not worthwhile because the total amount of
non-white PVC is small. The capacity of PVC reclamation is 2 tons per hour.
We are going to incorporate the granules into new window profiles by car-
rying out co-extrusion. The inner areas of the profiles then consist of recycled
material, whereas we use fresh material for the outer areas. The maximum per-
centage of recycled material is 80%.
We will incorporate the recycling PVC in new profiles first of all in our
plant in Sendenhorst.
Table 2 shows some properties of profiles made from virgin PVC compound
in comparison to the same profiles made of regenerated PVC covered with the
same virgin compound.
The results are mean values of 10 tests. All tested profiles perform accord-
ing to specification.
H. Uhlen 49
Table 2: Comparison of virgin and regenerated PVC
Regenerated
Test Method Unit Requirement Virgin PVC
PVC
DIN 53 753 2
Impact strength kJ/m 40 65 62
(double-V notch)
Impact strength by DIN-EN 477
o no break no break no break
falling weight (T=-20 C, m=1 kg)
Thermal shrinkage DIN-EN 479 % 2 1.6-1.7 1.4-1.6
Appearance after
DIN-EN 478 OK OK
heating at 150oC
Strength of welded
DIN-EN 514 N 5300 8000 7700
corners
CONCLUSIONS
Removed windows with plastic frames should not be brought to disposal
sites and they should not be incinerated. They should rather undergo reprocess-
ing and the reclaimed PVC in form of granules should be used for manufacturing
of the same product. The technical facilities for a fully automatic and clean recla-
mation of PVC are given. It is now up to authorities to assure that removed win-
dows are returned to recycling process in order to enable resource saving
production of new window profiles.
G. Voituron 51
Recycling PVC Bottles and Pipes

by Coextrusion
Georges Voituron
Solvay Research & Technology, Rue de Ransbeek 310, B-1120 Brussels, Belgium
INTRODUCTION
In our increasingly affluent society, concern for the environment as well as
for the quality of life is growing. A fraction of the public opinion is not far from
imagining that suppressing plastics would definitely solve the waste problem.
There is no doubt that mankind has to care for the future of this, his foster earth,
but this must be done scientifically and realistically, and not emotionally.
Nevertheless, in spite of their very important efforts towards a better envi-
ronment, efficient production to reduce energy consumption, significant reduc-
tion in pollution by using efficient modern affluent treatments, participation in
collecting and recycling of post-consumer and industrial waste, the chemical,
plastics, automotive, and other industries are often the scapegoats for this situ-
ation.
The only way for industry to change this situation is to inform as many peo-
ple as possible and as often as possible about its numerous environmental activi-
ties and achievements. That is the challenge that Solvay as polymers producer
and converter intends to achieve. The double goal to solve the plastics wastes
problem and save natural resources can be served by several strategies, for
which the E.C. has already set priorities. These priorities are as follows:
52 Recycling of PVC bottles and pipes by coextrusion
Recycling
• mechanical recycling
• item→item
• cascade: item to another item
Chemical recycling
• pyrolysis - partial or complete→fuel or gas
• hydrogenation→fuel
• dehydrochlorination
• complete pyrolysis→initial molecules →purification→polymerisation
Thermal recycling
• specific combustion (mono combustion: only plastics)
• clean (system)
• dehydrochlorination
Destruction in incinerators
• (Plastics in the municipal wastes)
• gas cleaning
• smoke
• energy recovering vapor electricity, hot water
Safe landfilling
Although, recycling is very popular, it cannot be achieved regardless of spe-
cific ecological and economic constraints. In particular, the viability of mechani-
cal recycling implies a subtle match between the available volumes of recyclable
materials, the quality and the cost of sorting and regeneration.
Some years of experience also clearly show that a sound policy must neces-
sarily combine several approaches, whereas in the field only well defined and re-
alistic objectives must be selected. This work describes the mechanical recycling
1
of PVC bottles and/or pipes to non-pressure coextruded three-layer pipes. In a
first step the material (bottles or pipes) must be regenerated and in the second
step, re-used. The coextrusion process also describes the modification of the pro-
duction equipment required by multilayer coextrusion.
G. Voituron 53
MECHANICAL RECYCLING
GENERAL APPROACHES
The mechanical recycling is recommended by authorities at a level of 10 to

15%. At the same time, the first R&D efforts have been committed to recovery of
single polymer species, with an option of their reprocessing either to the original
application or to less demanding use or preferably into durable items (for exam-
ple: bottles to pipes).
To fulfill this task, new expertise is needed in sorting, purification, repro-
cessing, and blending, in addition to sampling and evaluation methods.
The first set of operations is the preprocessing of items collected from mate-
rial recovery facilities, such as debaling, sorting, shredding, and coarse metal
elimination. Various operations can then be combined to achieve the first purifi-
cation stage: dry or wet grinding, wet centrifugation, and wet separation by den-
sity. These are followed by dry centrifugation, thermal drying, homogenizing
and micronization. In order to conduct these studies, state-of-the-art equipment
was purchased to evaluate continuous melt filtration.
According to the needs, new specialized modules will be added in future to
the existing basic equipment (for example, electrostatic sorting, in-line polymer
identification, etc.).
REGENERATION OF PVC PIPES
The collection of the recyclable pipes is organized by the producers associa-

tion. The pipes are delivered in bulk to the recycling plant. It is a mixture of PVC
and PE-pipes, with variable contamination by rubber, metals, sand, glass,
stones, etc.
The sequence of the purification operations is as follows (see Figure 1):
• manual pre-sorting of PVC and PE
• the pipes are then crushed and guillotined in a powerful press. In this oper-
ation the dimensions of the material are reduced to 30 cm to obtain better
handling characteristic. PVC pipes are broken whereas the more ductile
PE (if any remains) does not break, and can be easily identified and sorted
out in the following operations
• the material is sieved in order to remove the sand
• an operator sorts manually rubber and other non-PVC material from the
main PVC flow conveyed on a belt
Figure 1. PVC pipes regeneration.
• the material is ground, to reduce the particles size, and also to liberate the
admixtures which can be entangled; the material is sieved for the second
time to eliminate sand and other fine particles
• the material is purified from ferrous metals in a magnetic separator oper-
ating with an overband
• the material is then sieved in a classifier which accepts PVC powder be-
tween 1.5 mm and 15 mm
• after the last purification to remove very small particles of aluminum,
sand, and glass by shaking, the PVC powder is ground again until a parti-
cle size of less than 8 mm is reached, and finally micronised to 800 microns
• homogenization
REGENERATION OF PVC BOTTLES
The collection of bottles is organized by the “consortia” and operated by

subcontractors who also perform sorting. The sorted bottles are then baled and
sent to the regeneration section. After this treatment the material is ready to be
reused (second life). The sequence of the purification operations is as follows (see
Figure 2):
G. Voituron 55
Figure 2. Plastics recycling pilot plant.
• preparation of items to be recycled

• debaling in a drum with elimination of contaminants such as heavy materi-
als (metals, glass, stones, sand and also same caps and labels)
• sorting of the PET bottles out of the PVC bottles stream; various auto-
mated systems have been developed (x-ray, IR, NIR, polarized or UV light).
Such sophisticated equipments must however be backed up manually at
the final tuning stage, as they cannot deal with excessively crushed bottles and
bottle clusters. This brings us back to the collection level, where baling and
debaling techniques deserve some optimization.
The next step is the grinding of the bottles. This is a key point of the purifi-
cation because it is very important to control the size of the chips. In order to
minimize the amount of PVC fine particles which would be evacuated together
with paper pulp from labels and small mineral debris, through small holes (2.5
mm) of the centrifuge separators. Not only this PVC would be lost, but more im-
portant, its presence in the holes would disturb the elimination of the contami-
nants.
An in-depth investigation of the grinding operation has been done, to-
gether with a precise balance of all incoming material flow in the pilot line (see
the examples of a material balance and a diagram of the particle size optimiza-
tion).
The following parameters have been studied:
• dry and under water grinding, cold and hot water, water flow
• size of the holes in the grinder sieve
• speed of the grinder
• flow rate of the material.
After this purification, the material is submitted to another separation by
density (sink and float).
In this step, the light material (caps from PE or PP float) and the heavy ma-
terial (PVC but also PET sink). In the next operation, PVC is dried and stocked
in a silo homogenizer. At this stage of regeneration two possible ways existed,
the continuous melt filtration operated on a Knaus filter or the micronization.
We consider that the level of contamination of the PVC by the PET must be less
than 100 ppm. This level of purity can only be reached by an upstream auto-
matic system followed by a manual final tuning. Nowadays, PET flakes are of-
ten separated through “micronization” and sieving, based on the difference in
brittleness of both materials, with an exception for the non-bioriented parts of
PET bottles (neck and bottom), which micronize similar to PVC, thereby causing
residual contamination, as their melting temperature lies much higher than the
processing temperature of PVC. We consider this approach as insufficient.
As PVC-PET separation is an important step for both PVC and PET regen-
erators, research is being carried out in unconventional ways, i.e., electrostatic
separation (differential triboelectrification) and froth flotation are currently ex-
amined.
Aluminum caps are the only other contaminant. Although aluminum has
been widely replaced by PE, its separation still requires some development (e.g.,
hydrocyclones, Eddy currents, conductor/non-conductor electrostatic separa-
tion, or active metal detectors). The pilot plant is equipped with the Eddy cur-
rents separation.
G. Voituron 57
RE-USE OF PVC BOTTLES AND PIPES FOR COEXTRUDED PIPES

The recycling of the PVC bottles and pipes is performed in the same ma-
chines, only some technological details in the feedbox and in the head of the ma-
chine are different.
From the chemical point of view, the formulations of stabilization and lu-
brication differ because of the difference between the original stabilization of
bottles (calcium-zinc) and pipes (lead).
A market study showed that there is a very good opportunity to sell both
types of pipes in which PVC bottles and PVC pipes are recycled in the inner
layer. These pipes are already made and marketed under the following denomi-
nation:
• Renofort , large size, diameters 250, 315, 400, and 500 mm, three layer
pipes containing foamed recycled PVC bottles
• Renodur , small size, diameters 125 and 160 mm, three layer pipes contain-
ing rigid recycled PVC pipes.
DESCRIPTION OF THE COEXTRUSION PROCESS
A typical coextrusion line for pipes is shown Figure 3. It differs from classi-
cal PVC lines:
• the line includes two extruders
• a “feedblock” is introduced to merge the flows coming from the two extrud-
ers and to make a three layer annular parison (see Figure 4).
The feedblock is followed by a sizing device, designed to give the parison a
shape close to its final diameter and wall thickness. The rest of the line (calibra-
tion unit, cooling bath, saw, haul-off, socketing and packaging machines) does
not differ from conventional pipe extrusion lines.
FORMULATION OF THE PVC
THERMAL STABILITY
The formulation of all three layers (compact inner layer, foamed or compact
intermediate layer, and compact outer layer) must take into account that be-
cause of the length of feedblock, the polymer average residence times are much
longer in coextrusion than in conventional pipe extrusion. Thus, thermal stabil-
ity has to be increased.
Figure 3. PVC pipes recycling.
Figure 4. Feed-block.
G. Voituron 59
LUBRICATION
The pressure drop in the feedblock is important because the balance be-
tween the fluxes for the three layers is achieved by a strong annular restriction
in the flow channel. This restriction is similar to a chocker bar on a flat die. Lu-
brication (external and internal) is used to reduce the pressure drop in the
feedblock-die assembly and hence, to reduce the outlet pressure at the extruder.
Adjustment of the lubrication is necessary to reduce the pressure drop in
the feedblock and in the die. Optimization of the lubrication level is done by ex-
truding different formulations through two dies on a laboratory scale extruder.
The first die has smooth walls and the other, grooved walls. In the first die, be-
cause of lubrication at the interface with the metal, the polymer is allowed to slip
as is the case in the feedblock and die used for tube coextrusion. In the second
die, the grooves prevent slipping. Comparison between the behavior of the two
dies (with and without slipping) allows for the calculation of the slip velocity.
The influence of slip velocity on feedblock and die pressure drop has been mea-
sured experimentally and confirmed by numerical flow simulations. After up-
grading, only minor differences in slip velocity remain between different
batches.
IMPACT RESISTANCE
Recycled PVC separation has shown the presence of up to 8% impact modi-

fier. Impact resistance tests, performed with an instrumented falling weight,
have shown that this level of impact modifier is more than adequate: coextruded
pipes perform better under impact when the foamed layer is made of recycled
PVC compared to a reference formulation based on virgin resin.
EXPANDABILITY AND FOAMING PROCESS
The acrylic modifiers already present in the recycled PVC have a positive
effect on stretching behavior and no further modification needs to be made. The
gelation level has a major influence on the draw behavior of the hot PVC melt.
The influence of the formulation on the gelation has been measured.
Information have been collected regarding gelation time and maximum
torque for some formulations. One of them have been chosen as the best compro-
mise, i.e., it provides sufficient lubrication to reduce pressure drop in the
feedblock and the die, while allowing for sufficient gelation. The foam is pro-
duced with chemical blowing agents. Azodicarbonamide is often used with PVC.
The influence, of the blowing agent on the foam density at different tempera-
3
tures is an important factor. A foam density of 0.7-0.8 kg/dm is a good compro-
mise between weight reduction and impact resistance. The chosen blowing
agent gives a good density level and a low sensitivity to temperature.
FLOW STABILITY
Finally, we have seen that a low extrusion temperature was needed to con-
trol the viscosity of the foamed layer and maintain stability of flow for the
multilayer structure. Material temperature after 10 min of residence time in the
Brabender Plasticorder is used as an indicator of the melt temperature attained
in the extruder. Here again, the formulation chosen to maximize the
expandability gives a good compromise. Lubrication of the external compact lay-
ers could be used as an alternate means to reduce flow unstability, but the addi-
tion of lubricant is limited by the impact resistance.
MODIFICATION OF THE PRODUCTION EQUIPMENT
Some modifications of the standard production line of coextruded pipes

have been necessary to process the PVC bottle scraps under optimal conditions.
The scraps entering the factory are first homogenized in a blending silo to
reduce material fluctuations due to variation in bulk density and in scrap for-
mulation.
Screw profiles were adapted to the control of the extrusion temperature.
The temperature increase along the screws have been calculated for three differ-
ent profiles. Screw 2 of the Figure 5 has been chosen for production.
On counter-rotating screw extruders, screw tips are known to cause hot
spots in the melt. This problem has been investigated. Since the temperature
has the major effect on the foaming process, this situation can result in cir-
cumferential heterogeneity of the parison. Thermal conditioning of the screw
tips or appropriate lubrication, reducing the temperature increase of the melt,
can solve the problem.
Finally, the length of cooling baths has to be increased to prevent ovality of
the pipe: indeed, the foamed layer acts as an insulator, and reduces the effective-
ness of the cooling units.
G. Voituron 61
Figure 5. Calculated temperature profile along different screws.
PRODUCT EVALUATION
The pipes produced with a foamed layer made of recycled PVC satisfy the
requirements of ISO/DP 9970 for building discharge pipes and ISO/DP 9971 for
sewage pipes.
For building applications, diameters for soil, waste, drainage range from
32 mm to 200 mm with a wall thickness of about 3 mm. Sewage pipes have a di-
2
ameter between 125 and 500 mm; two different stiffness classes, 4 kN/m and 8
2
kN/m are used; wall thickness varies from 3 mm (125 mm diameter, stiffness 4
2 2
kN/m ) to 15 mm (500 mm diameter, stiffness 8 kN/m ). Impact resistance satis-
fies ISO standard 3127 for compact PVC pipes. Socketing of the pipe or separate
fittings are provided.
CONCLUSIONS
Coextruded light-weight pipes having an intermediate foamed layer have
been on the market for some time. This product has proven advantages over
compact single layer pipes or traditional pipes (clay or concrete):
• good mechanical properties
• good resistance to flattening due to soil loading or traffic
• long life
• high chemical inertia
• good impact resistance

• ease of handling and installing
• compatibility with existing complete fittings programs
• ease of cutting and of connecting to existing pipe systems
• flexibility in the number and locations of connections to houses.
In this paper, we have shown that recycled PVC bottles can be processed to
form a foamed layer of coextruded light-weight pipes. Properties of the pipes
produced with recycled PVC are at least as good as those achieved when foamed
layer consists of virgin PVC. The pipes with recycled PVC meet the technical
standards for building discharge pipes and sewage pipes, including stiffness
and impact resistance requirements.
The foamed layer of light-weight pipes offers a substantial outlet for recy-
cled PVC bottle. However, commercial viability of the project is still hampered
by two problems:
• the availability and cost of PVC bottles: collecting of post-consumer scraps
from households is still very limited. Only a wide involvement of public au-
thorities could lead to an increased availability and reasonable price. Recy-
cled PVC from industrial sources is readily available (several thousands
ktons/year) and is already used by the plastics processors
• the lack of normalization in the technical specifications for light-weight
pipes; for example, in the Netherlands, technical standards do not allow for
the use of recycled material.
REFERENCE
1. C. Dehennau, S. Dupont, P. Benjamin, B. Rijprema, and G. Voituron, Recycle’91,
Davos, April 3-5, 1991.
F. P. La Mantia 63
Basic Concepts on the Recycling of

Homogeneous and Heterogeneous Plastics

Dipartimento di Ingegneria Chimica dei Processi e dei Materiali,
Università di Palermo, Viale delle Scienze, 90128 Palermo, Italy
The main problems in post-consumer plastics recycling are due to the degradation of polymers during
processing steps and during lifetime. In the recycling of heterogeneous waste, the incompatibility
among the different phases also poses substantial difficulty. The influence of both degradation and in-
compatibility on the structure and on the properties of the recycled “secondary materials” is shortly re-
viewed.
INTRODUCTION
Plastics recycling is a very recent technology, or rather a group of technologies,
1-4
which has not yet found firm base. In effect, the term recycling is generally
taken to mean a series of processing operations carried out on plastics to produce
“secondary materials” for manufacturing of various types of articles.
Post-consumer plastics recycling is not only a high-priority item for today’s
public opinion, it has become an urgent necessity. It is obvious that simple recy-
cling to produce “secondary materials” cannot by itself solve the environmental
problems caused by discarded plastics not only because of large quantities in-
volved, but also because of the many types of polymers difficult to separate or
process together. In the latter case, in particular, articles produced with the re-
cycled material would have poor mechanical properties and few possibilities of
application.
In this paper, the main problems arising during recycling of post-consumer
plastic materials are briefly discussed in order to form some basic concepts on
64 Basic concepts on the recycling
these processing operations. The discussion concentrates on the melt processing

operations without any considerations about the preliminary operations such as
washing, drying, etc.
MAIN PROBLEMS IN PLASTICS RECYCLING
Post-consumer plastics recycling is complicated by two main problems:
• degradation during processing and lifetime (thermomechanical degrada-
tion, photooxidation, etc.)
• incompatibility between different polymers.
This latter point is only important for recycling of mixed heterogeneous
plastics.
Other problems can arise from:
• different melting points of the polymers in mixed plastics
• different size and shape of the materials
• low apparent density.
The degradation during lifetime is mainly caused by photooxidation pro-
cesses. The changes in the structure of the polymers are similar to those induced
by thermomechanical degradation and differing by reactions with oxygen, in
general not present in the processing machines, leading to the formation of oxy-
genated groups in the macromolecular chain. The presence of different poly-
meric materials can remarkably increase the difficulty of recycling operations.
In particular, different melting points can induce degradation of some compo-
nents of the mixtures, whereas the incompatibility among the different phases
can give rise to materials with poor mechanical characteristics.
In the following, the influence of the degradation on structure and proper-
ties of the polymeric materials will be reviewed, considering in particular the
thermomechanical degradation. As for the blends, the basic concepts regarding
the behavior of polyphasic systems and compatibilization will be presented.
DEGRADATION
During the processing and lifetime of the plastic articles, heat, mechanical
stress, and ultraviolet radiation can deeply change the structure and the mor-
phology of polymer. Although photooxidation and thermomechanical degrada-
tion induce variable changes in the structure of the plastic materials, it is
possible to categorize the effects of these degradative processes as:
• change in molecular weight and molecular weight distribution
F. P. La Mantia 65
• formation of chain branching

• formation of oxygenated compounds, unsaturations, etc.
It is worth noticing that the change of molecular weight can induce also
changes in the crystallinity.
The degree and type of degradation depend on the processing conditions
and on the nature of polymer, but all changes to which chemical structure was
subjected are followed by the changes in the properties of the material. Most pro-
cessing operations use fairly mild processing parameters and polymers of mod-
erate molecular weight: under these conditions degradation is, generally,
modest. More severe processing conditions, polymers with high molecular
weight, or repeated processing operations can cause significant decrease in
polymer characteristics.
As an example, Figure 1 shows the change of molecular weight and elonga-
tion at break versus the number of extrusions of polyethyleneterephthalate
5
taken from carbonate beverage bottles. Figure 2 reports the elongation at break
of a polyethylene, taken from greenhouse film, as a function of the
6
photooxidation time.
Figure 1. Molecular weight and elongation at break vs. number of extrusions of PET sample.
Figure 2. Elongation at break vs. photooxidation time for PE sample.
The reduction of molecular weight and elongation at break of PET sub-

jected to several extrusions processes is rapid in the first steps (first recycling),
and it becomes less pronounced during the following recycling steps. The molec-
ular weight and then the viscosity of the recycled material does not allow for its
use in blow-molding operations, where high viscosity is required, and gives rise
to a significant reduction of the elongation at break.
The influence of the photooxidative degradation on the elongation at break
of polymers is also very dramatic. This effect has been attributed to the degrada-
tion of the macromolecules but also to the formation of physical and chemical de-
6
fects on the surface.
INFLUENCE OF STRUCTURE AND MORPHOLOGY
ON THE BEHAVIOR OF POLYMERS
The changes in the structure of polymers during degradation processes, pro-
duce, in their turn, changes in the properties of these materials. Each modifica-
tion influences differently the behavior of the polymeric material. Indeed,
changes of molecular weight, Mw, molecular weight distribution, MWD, and for-
mation of long and short chain branching induce remarkable changes in the
melt properties of polymers and smaller variations in the solid properties. On
the contrary, the formation of long and short chain branches and the increase of
crystallinity dramatically influence the solid properties. Finally, the formation
of new chemical bonds and new chemical groups (in particular double bonds and
oxygenated groups) strongly influence the chemical and physical resistance and
F. P. La Mantia 67
in particular the photooxidation behavior. All these considerations are schemat-

ically reported in Table 1.
Table 1: Influence of the structure and morphology on rheological
and mechanical properties and photoxidation resistance
Properties
Parameter
rheological mechanical photooxidation resistance
Mw, MWD PP P P
Branching (long) PP PP PP
Branching (short) P P PP
New chemical groups − − PP
Crystallinity − PP P
PP strongly influenced, P influenced, − not influenced or unpredictible
The influence of Mw, MWD, and branching, which can remarkably change
7,8
during recycling operations, on the rheological and mechanical properties of
the “secondary materials”, is discussed below.
MOLECULAR WEIGHT
The molecular weight has a strong influence on the rheological behavior of poly-
meric systems. This means that small changes in molecular weight can deeply
change the viscosity of molten polymer. It is worth noticing, however, that in
most processing operations, the variations of molecular weight induced by
thermomechanical stress are very small, as a consequence only the Newtonian
viscosity or the viscosity at low shear rates is strongly influenced by the molecu-
lar weight. In Figure 3, the flow curves of a polypropylene sample recycled sev-
eral times by injection molding are reported. It is evident that while the
Newtonian viscosity or the viscosity at low shear rates drastically decreases
with the number of recycling steps, (i.e., increasing the amount of degradation
and decreasing the molecular weight), the viscosity at high shear rates is, on the
contrary, less affected by the processing. In particular, in the shear rate range
-1
typical for the industrial processing operations (100-1000 s ), only a change of a
factor of about five in the viscosity occurs whereas the Newtonian viscosity dif-
fers by about two decades. This means, that during usual processing operations,
small changes of molecular weight, would cause no significant variations of the

viscosity and the extrusion pressure or torque. The above example refers to poly-
mers which undergo chain scission during melt processing. However,
thermomechanical stress can also induce increase of molecular weight and then
9
viscosity. This behavior depends on the nature of the polymer. For example,
HDPE mostly undergoes branching and crosslinking with increase of Mw, while
PP does not. Also, the processing conditions limit changes in polymer. At high
temperatures, for example, high density polyethylene shows a small but signifi-
cant reduction of viscosity.
Figure 3. Flow curves of PP samples; virgin (TQ), and recycled one (S1), two (S2), three (S3), and
four (S4) times.
In other types of flow, in particular in elongational flow, the influence of the

molecular weight of the molten polymer may be remarkable. A decrease of the
molecular weight (decrease of polymer viscosity) can induce a smaller resistance
of melt during drawing in fiber spinning, film blowing, and blow-molding. In
Figure 4 the melt strength, i.e., the force sustained by the melt during drawing
(non-isothermal elongational flow) is plotted as a function of the molecular
10
weight for a sample of low density polyethylene. A decrease of the molecular
weight gives rise to a decrease of melt strength; the polymer cannot sustain the
F. P. La Mantia 69
stress of drawing and its own weight, and the melt breaks. The formation of the
products becomes impossible. This behavior suggests that in most cases the re-
cycled polymers cannot be processed in the same processing operations used for
the virgin material.
Figure 4. Melt strength as a function of the molecular weight.
BRANCHING
The formation of short and long chain branching can occur during melt process-
ing of polymers as a result of chain scission and recombination of radicals. The
polyethylenes are the typical polymers which can undergo this structural modi-
fication. As said previously, long chain branching strongly influences the rheo-
logical and mechanical properties, whereas the presence of short branches
influences mainly the solid properties. Both types of branches worsen the
photooxidation resistance.
The presence of long chain branching decreases viscosity, and in particular
the Newtonian viscosity, compared to linear polymers with about the same mo-
lecular weight, and, in general, gives rise to a more pronounced non-Newtonian
behavior and instability phenomena (melt fracture) at lower output flow rates.
As far as the mechanical properties are concerned, both short and long chain
branching, causing a dramatic reduction of crystallinity, give rise to a drastic
decrease of these characteristics. However, this feature is more remarkable
with the long chain branching. This behavior is evidenced in Table 2 where
modulus and tensile strength of three samples of polyethylene (linear and hav-
ing either short or long chain branching) of comparable molecular weight, are
reported.
Table 2: Mechanical properties of polyethylenes

with different degrees of branching
Sample Modulus (MPa) Tensile strength (MPa)
HDPE 400 32
LLDPE 220 25
LDPE 70 15
Figure 5. Typical property-composition curves.
The presence of branches leads to reduced crystallinity and then a decrease

in the rigidity of the material. On the other hand, elongation at break and im-
pact strength can improve.
POLYMER BLENDS
It is a well-known fact that only a few polymer pairs are compatible. Moreover,
the concept of compatibility is in itself difficult to define and quantify. Although,
it is not completely correct, for our purposes, we can define the compatibility on
the basis of the property-composition curve, where properties are taken as mac-
roscopic characteristics such as mechanical or rheological properties, etc. Figure
5 plots the qualitative trends of property-composition curves for blends which
F. P. La Mantia 71
are compatible, partially compatible or incompatible. The synergistic effect

(maximum in this curve) is found only in blends with strong interactions be-
tween the two phases. Instead, the antagonistic effect is typical of those polymer
pairs with strong repulsion (for example, apolar polymers and strongly polar
polymers). In most cases, the properties of blends are intermediate between
those of the two components, (partially compatible or semi-compatible), al-
though, most of them show values lower than expected on the basis of an addi-
tive law. More over, for the same blend not all properties behave in the same
11
way. In Figure 6, we can see how, for a typical incompatible blend,
LDPE/polyamide-6, the tensile strength values are intermediate between those
of homopolymers, while elongation at break shows a deep minimum. The mor-
phology of this blend, SEM micrograph of Figure 7, suggests that the poor me-
chanical properties are due to the very scarce adhesion between the two phases.
Figure 6. Tensile strength and elongation at break vs. polyamide content in LDPE/polyamide-6
blends.
A third component may be added to the blend in a small amount to act as a
bonding agent between the two incompatible phases. This macromolecular com-
pound is, in most cases, a copolymer formed from the monomers of the two
homopolymers making up the mixture. The copolymer is thus miscible in the
two phases, creating a bond between the two completely immiscible phases.
Adding these compatibilizing agents produces an effect on several mechanical
12
properties which can be clearly seen in Figure 8, where the addition of poly-
propylene functionalized with maleic anhydride, PPmal, increases the elonga-
tion at break of incompatible polypropylene/polyamide-6 mixtures. In this case,
there is not only a physical bond produced by the compatibilizing agent, but also
a chemical bond, since the carbonyl groups in the functionalized polypropylene
react with the amine groups of polyamide, anchoring the two phases even more
firmly. The improvement of adhesion is evident from SEM micrographs of the
PP/Ny blend with and without compatibilizer, Figure 9 a and b, respectively.
Figure 7. SEM micrograph of a LDPE/Ny-6 blend with 10% Ny.

F. P. La Mantia 73
Figure 8. Elongation at break vs polyamide content for PP/polyamide-6 blends.
Lastly, we must bear in mind that even in the presence of an excellent

compatibilizing agent, it is not possible to improve all the properties of the mix-
ture with respect to properties of both components, and in some cases, it is im-
possible to avoid the minimum in some property-composition curves.
Incompatibility and antagonistic effects can arise not only from blending
polymers with different structures, but also from blending polymers with very
similar structures. This can occur when blends are prepared from the same poly-
mer (virgin and recycled).
13
Figure 10, shows the curves of elongation at break as a function of the
amount of virgin polypropylene blended with several degraded polypropylene
samples. The elongation at break-composition curves show a significant mini-
mum whose depth and amplitude depend on the characteristics of the degraded
material. High elongation at break may be obtained only with very high percent-
ages of virgin material. These blends show then a strong incompatible behavior
that has been attributed to the different crystalline structure of the two compo-
nents. Polypropylenes of similar molecular weight show well-interconnected ag-
gregates of similar size. As degradation increases, the recycled polypropylene
with a lower molecular weight and increased number of oxygenated groups,
crystallizes into aggregates of larger sizes. This causes a heterogeneous struc-

ture with weak interconnections, leading to modest elongation at break val-
13
ues.
Figure 9. SEM micrographs of a PP/Ny blend without (a) and (b) with compatibilizer.
Generally speaking, it is not possible to predict the physical properties of a

mixture containing recycled material. Consequently, when degraded polymers
are used, it is a good idea to add only limited amount to the virgin material if the
mechanical properties of virgin polymer are to be maintained.
F. P. La Mantia 75
Figure 10. Elongation at break vs. recycled PP content for PP/PPdeg blends. The degradation de-
gree increases from £, , l, to n.
CONCLUSIONS
• The recycling of homogeneous post-consumer plastics presents many prob-
lems mainly because of degradation of the polymeric materials during pro-
cessing and lifetime. The degradation processes, which can become very
important in recycled materials subjected to many repeated processing op-
erations, lead to drastic reductions of the mechanical properties of the sec-
ondary materials.
• In the case of recycling of heterogeneous plastics, the situation is still more
complicated mainly due to the incompatibility between different polymeric
phases. The properties of the recycled materials are in general poor and can
not be predicted only on the basis of the properties of individual compo-
nents. Improvements of the mechanical properties can be achieved by add-

ing compatibilizing agents.
• In the case of blends made with the same polymer (virgin and recycled), the
incompatible blends can result and only small amounts of recycled mate-
rial can be used to avoid drastic decrease in the mechanical properties.
ACKNOWLEDGMENT
This work has been financially supported by MURST.
REFERENCES
1. J. Leidner in Plastics Waste, M. Dekker, New York, 1981.
2. F. P. La Mantia in Recycling of Plastic Materials, Ed. F. P. La Mantia, ChemTec,
Toronto, 1993.
3. G. D. Andrews, P. M. Subramanian, Eds, in Emerging Technologies in Plastic
Recycling, ACS, Washington, 1992.
4. F. P. La Mantia, Macplas International, May 1990, p. 53.
5. F. P. La Mantia and M. Vinci, Polymer Recycling, 1(1994).
6. F. P. La Mantia, Europ. Polym. J., 20, 993 (1984).
7. F. P. La Mantia, International Conference on Advances in the Stabilization
and Controlled Degradation of Polymers, Luzern, 25-27 May, 1988.
8. F. P. La Mantia, Materials Engeneering, 2, 351 (1991).
9. F. P. La Mantia, A Valenza, and D. Acierno, Polym. Deg. Stab., 15, 283 (1985).
10. F. P. La Mantia and D. Acierno, Polym. Eng. Sci., 25, 279 (1985).
11. F. P. La Mantia and A. Valenza, Europ. Polym. J., 25, 553 (1989).
12. F. P. La Mantia, Adv. Polym. Tech., 12, 47 (1993).
13. A. Valenza and F. P. La Mantia, Polym. Deg. Stab., 20, 63 (1988).
B. Ballauri, M. Trabuio, and F. P. La Mantia 77
Compatibilization of Recycled
Polyethyleneterephthalate/Polypropylene
Blends Using a Functionalized Rubber
B. Ballauri
F.lli Maris SpA, 10090 Rosta (TO), Italy
M. Trabuio
Shell Italia, SpA, Via Chiese 74, 20126 Milano, Italy
F. P. La Mantia
Università di Palermo, Viale delle Scienze, 90128, Italy
The recycling of heterogeneous plastic waste gives, in general, a secondary materials with poor proper-
ties because of a scarce compatibility among the polymers present in the waste. Compatibilization of
heterogeneous blends is thus a primary aim of the research on recycling. Blends of polyolefins and
polyesters are particularly difficult to design due to the different chemical nature of the two classes of
polymers. In this work the goal is achieved by using a commercial maleic anhydride grafted SEBS rub-
ber. Blends of recycled polypropylene and polyethyleneterephthalate show a morphology with the
large voids, due to a limited adhesion between the two phases, and a brittle behavior with poor me-
chanical properties. By using small amounts of this functionalized rubber, the two phases become al-
most indistinguishable and some mechanical properties are also considerably improved. Tensile
strength and thermomechanical resistance are only marginally lower.
78 Compatibilization of recycled PET/PP blends
INTRODUCTION
Polymer blends offer the possibility to prepare polymeric materials with proper-
ties tailored to specific needs. In plastics recycling, blending of different poly-
mers is in some cases necessary because the separation of a single component
can be difficult or expensive. In most cases, however, the polymers are immisci-
ble and this leads to blends having poor properties.
Compatibilization through a third component, acting as a bridge between
the two incompatible phases, becomes necessary to stabilize the blend and to im-
prove its mechanical performance.
Copolymers synthesized from monomers miscible with two components of
the blends or formed by in situ reaction during processing reduce the interfacial
tension between the two phases.
Polyolefins, polyethylene and polypropylene, PE and PP, in particular, and
copolyesters, mostly polyethyleneterephthalate, PET, are frequently encoun-
tered in urban and industrial plastics waste and are recycled after separating
the polymers by flotation. It is true that the separation is quite easy and not ex-
pensive, but the heterogeneous recycling of these two classes of thermoplastics
can still lead to the obvious advantages. Beside the lower cost of the recycling,
the blends can give rise to secondary materials with interesting properties. In
particular, blends of polyolefins and PET can display good mechanical and per-
1
meation characteristics.
In order to obtain materials stable with good properties, the blends have to
1-7
be compatibilized. In this work, we have investigated the effects of the
compatibilization on the structure, rheological, and mechanical properties of
blends of PET and PP. The compatibilizer used in this study is a triblock copoly-
mer consisting of polystyrene end-blocks and poly(ethylene-butylene)
mid-blocks grafted with maleic anhydride, MA-g-SEBS. This copolymer was al-
5,8,9
ready used to compatibilize other blends of polar and apolar polymers with
satisfactory results.
EXPERIMENTAL
MATERIALS
The materials used in this work were two recycled thermoplastics (PET and PP)
and a virgin elastomer. PET has been obtained from a separate collection of
post-consumer bottles and has an average intrinsic viscosity of about 0.7 dl/g.
PP has been recycled from industrial scraps. The melt flow index was about 5.5.
Kraton FG 1901X is a triblock copolymer consisting of polystyrene end-blocks
and a partly hydrogenated polybutadiene mid-blocks grafted with maleic anhy-
dride, and manufactured by Shell Chemicals.
Two blends were reprocessed, namely PET/PP 1 with equal amounts of the
two polymers and PET/PP 2 where the content of PET is two times that of PP.
The concentration of Kraton ranged from 0 to 10%.
RECYCLING
Recycling of the blends was carried out using a laboratory corotating twin screw
extruder R&D30 manufactured by Maris, Italy.
The thermal profile used for all the extrusions was 200-250
o o
270-285-285-260-250 C. The die temperature was 250 C, and the screws speed
ranged from 150 to 500 rpm. This thermal profile was chosen to induce a rapid
melting and degassing of the material and then to reduce the temperature to
minimize the degradation of PET. In fact, the temperature of the melt ranged
o
from 255 to 265 C. The torque was recorded during extrusion.
The specimens for the mechanical testing were prepared by injection mold-
o
ing using a Boy T 50 machine. The melt temperature was about 285 C and the
o o
mold temperature about 20 C. The materials were dried under vacuum at 110 C
for 2 h before molding.
STRUCTURAL DETERMINATIONS
Thermograms have been recorded using a Perkin-Elmer DSC 7 at a heating rate

o
of 10 C/min.
SEM micrographs have been obtained using a Jeol microscope model T 300.
The fractured samples were gold sputtered to make them electrically conduc-
tive. Image analysis was performed using an optical microscope Leica connected
with an image analyzer Quantimed 500.
RHEOLOGICAL AND MECHANICAL PROPERTIES
Flow curves of the pure polymers were measured with a capillary viscometer
o
Rheoscope (CEAST, Italy) at 270 C with a capillary D=1 mm and length to diam-
eter ratio of 40.
Rheological properties of the blends were continuously monitored by
means of an on-line Gottfert Real Time Rheometer installed on the extruder
Figure 1. Thermogram of the PET/PP 1 blend.
head. The capillary die had a diameter of 4 mm and a length-to-diameter ratio of

15. Rabinovitch correction was applied throughout.
Stress-strain curves have been determined by using an Instron machine at
-1
an elongational rate of 0.8 min .
Impact strength was determined using a Fractoscope, CEAST.
All the results are average of at least seven measurements.
Vicat tests were performed under a weight of 10 N.
RESULTS AND DISCUSSION
STRUCTURAL CHARACTERIZATION
The thermogram of the blend PET/PP 1 is reported in Figure 1. Three peaks are
displayed, namely: the melting points of PP and PET and the crystallization
o
temperature of PET at about 130 C. The two melting points are located at the
same temperatures as those of the pure polymers. This indicates that no interac-
tion occurs in crystalline phase. The glass transition temperature located at
o
about 80 C is the same as Tg of pure PET, meaning that the two phases do not in-
teract also in amorphous phase.
The incompatibility of the two polymers is confirmed by the morphology of
the blend PET/PP 1 as revealed by the SEM micrograph of Figure 2. The contin-
uous phase is composed of polypropylene and the discrete particles of PET. This
has been established by X-ray energy dispersion analysis microprobe and was
expected on the basis of the viscosity of the two components in the processing
conditions. In the case of blends with equal amounts of the two components, the
10
low viscosity component forms the discrete phase.
Figure 2. SEM micrograph of the PET/PP 1 blend.
o
The flow curves at 270 C of PET and PP are plotted in Figure 3. The tem-
perature has been chosen considering that the die temperature is lower than the
melting point of PET and that the morphology of the blend is probably the same
o
as that formed at 270 C (the temperature at which the melting of the PET is
complete) (see Figure 1).
-1
At the shear rate in the extruder (100-200 s ), PET shows viscosity lower
than PP which then becomes the dispersed phase. The two phases show a very
poor adhesion and the dispersed PET particles, having average dimensions of
about 1-10 µm, are completely separated by the continuous PP matrix.
Figure 3. Flow curves of pure components.
By adding small amounts of Kraton, the morphology of the blend is remark-

ably modified. The micrographs of the blends with 2.5 and 10% are reported in
Figures 4 a and b, respectively. Contrary to Figure 2 (unstabilized blend), the
pictures of the compatibilized blends show that the two phases adhere very well,
even at low content of Kraton, and the dimensions of the discrete particles are
very small. These features, and in particular the reduction of the dimensions of
the PET particles, are only slightly enhanced with increasing the content of
Kraton. The particle size distribution of the samples with 0, 2.5, and 10% of
Kraton are reported in Figures 5 a-c, respectively. As qualitatively seen from
SEM micrographs, the particle dimension is strongly reduced by adding the

functionalized rubber. Moreover, the particle size distribution of the
unstabilized blend is approximately a Gaussian curve, whereas the distribution
of the compatibilized blends has maximum in the low diameter zone with only a
few larger particle sizes present. Both the average diameter and the particle size
distribution then decrease and this behavior is remarkable also at low concen-
tration of Kraton, as displayed in Figure 6, where the average diameter of the
particles is plotted as a function of the content of the functionalized rubber. Al-
though, the particle size and distribution decrease with increasing the content
of Kraton, the reduction occurs mainly at low content of the compatibilizer. With
Figure 4. SEM micrographs of the PET/PP 1 blend with 2.5 % Kraton (a) and 10% Kraton (b).
Figure 5. Particles size distribution.
Figure 6. Average diameter of the PET particles in the PET/PP 1 blend as a function of the Kraton
concentration.
Figure 7. Scheme of the reactions between MA-g-SEBS and PET.
the Kraton content increasing, the curves flatten and only small gain in the size
reduction results from higher Kraton concentrations. Similar results have been
obtained for the PET/PP 2 blend.
All structural determinations confirm the compatibilization action of the
functionalized MA-g-SEBS. A possible mechanism is reported in Figure 7. The
PET-MA-g-SEBS copolymers formed during processing act as a bridge between
the two phases improving adhesion of the two polymers and the properties of the
blend.
RHEOLOGICAL PROPERTIES AND PROCESSABILITY
The torque during extrusion of the two blends is reported in Figure 8 as a func-
tion of Kraton content. The torque increases with the concentration of Kraton
but this increase is rather small. The increase of the torque, especially at high
content of the functionalized rubber, cannot be attributed to the presence of

Kraton because of its low viscosity compared with two other components (see
Figure 3). The increase of the energy necessary to extrude the blend must be at-
tributed to the compatibilization induced by the functionalized rubber. The vis-
cosity, indeed, increases with the concentration of Kraton as reported in Figure
9 and 10 for PET/PP 1 and PET/PP 2, respectively. Because of the biphasic na-
ture of the blends, the viscosity has been plotted against the shear stress, the
stress is the same in both phases while the shear rate changes in the phases. The
growth of the viscosity is large at low shear stress whereas, at high shear stress,
the viscosity is only slightly dependent on the Kraton content. This means, as al-
ready seen with the torque data, that the processability, carried out at high
shear rates, is only slightly worsened by adding the compatibilizing rubber. It is
worth noticing that the increase of the viscosity mostly occurs at low concentra-
tion of Kraton; the viscosity-concentration curves flatten with increasing the
rubber content. This feature, which confirms the morphological data above re-
ported, indicates, similar to theoretical models, that the compatibilizers work at
low concentration, and the further addition of compatibilizer does not affect
morphology and properties.
Figure 8. Torque during extrusion as a function of the Kraton concentration.

Figure 9. Viscosity at fixed shear stress for the PET/PP 1 blend as a function of the Kraton concen-
tration.
Figure 10. Viscosity at fixed shear stress for the PET/PP 2 blend as a function of the Kraton con-
centration.
MECHANICAL PROPERTIES
Elongation at break, and impact strength are reported in Figures 11-12, respec-
tively for both blends as a function of the Kraton concentration. For both blends
the tensile strength (here not reported) is slightly decreasing with increasing
the content of Kraton. Elongation at break and impact strength, on the contrary,
increase in impressive way with increasing the content of the functionalized
rubber. With 5% of Kraton, the impact strength quadruples compared with
non-compatibilized blends and the elongation at break is ten times larger. The
stress-strain curves of the non-compatibilized PET/PP 1 blend and of the same
blend with 5% of Kraton, Figure 13, clearly highlight that the presence of the
functionalized rubber induces brittle-ductile transition yielding ductile mate-
rial. The decrease of the tensile strength is expected because the rubber shows
low values of these ultimate properties, whereas the impressive increase of elon-
gation and impact strength at break can be attributed to the properties of elasto-
mer and the compatibilization induced by the Kraton. While at high content of
the rubber the first explanation holds, at low concentration of Kraton, the
compatibilization must be considered responsible for the enhanced values of
elongation at break.
Figure 11. Elongation at break as a function of the Kraton concentration.

Figure 12. Impact strength as a function of the Kraton concentration.
Figure 13. Stress-strain curves of PET/PP 1 blend and of the same blend with 5% Kraton.
Figure 14. Vicat temperature as a function of the Kraton concentration.
Figure 15. Impact strength for PET/PP 1 blend with 5% Kraton as a function of screw speed.
The presence of the rubber can give rise to some worsening of the thermal
resistance of the blends. To verify this behavior, tests for measuring the Vicat
temperature have been carried out. In Figure 14, the Vicat temperature is re-
ported as a function of the content of Kraton. As expected a reduction of the Vicat
temperature of the non-compatibilized blend is observed but can be considered
still reasonable: with 5% of Kraton the Vicat temperature decreases by about
o
6-7 C for both blends.
Because the properties of the blends depend on the morphology, deter-
mined by mixing conditions and in particular by the stress applied to melt, the
effect of the mixing speed was investigated for the PET/PP 1 blend including 5%
Kraton. Most mechanical properties are only slightly dependent on the screw
speed, but impact strength is considerably affected. In Figure 15, impact
strength growth rapidly at low screw speed, then with a further speed increase
its effect is diminishing.
CONCLUSIONS
Blends of PET and PP (and in general blends of polyesters and polyolefins) re-
sult in materials with inferior mechanical properties because of the incompati-
bility between the two phases. By adding a rubber functionalized with maleic
anhydride, the morphology and the properties of the blends are strongly modi-
fied and the resulting materials show good mechanical properties. In particular,
small amounts of this compatibilizer induce a brittle-to-ductile transition with
an impressive improvement of some properties, such as, elongation at break and
impact strength. SEM analysis allowed to reveal the compatibilizing action of
the functionalized rubber in these PET/PP blends.
REFERENCES
1. P. Bataille, S. Boissè, and H. P. Schreiber, Polym. Eng. Sci., 27, 622 (1987).
2. T. D. Traugott, J. W. Barlow, and D. R. Paul, J. Appl. Polym. Sci., 28, 2947, (1983).
3. In-Mau Chen and Chii-Ming Shiah, Plastics Engineering, October 1989, p. 33.
4. J. Curry and A. Kiani, ANTEC ‘90, p. 1452.
5. F. P. La Mantia and M. Vinci, Polymer Recycling, 1, 35 (1994) .
6. G. E. Wissler, ANTEC ‘90, p. 1434.
7. P. Sambaru and S. A. Jabarin, ANTEC ‘92, p. 586.
8. R. Holsti-Miettinen and J. Seppala, Polym. Eng. Sci., 32, 868, (1992).
9. C. Boman Ducommun, Recycle ‘91, p. D5.3.
10. L. A. Utracki in Polymer and Alloys and Blends, Hanser Verlag, Munich, 1989.
F. P. La Mantia 93
Recycling of Post-consumer
Plastic Containers for Liquids
Francesco P. La Mantia
Università di Palermo, Viale delle Scienze, 90128 Palermo, Italy
Mechanical properties of secondary plastic materials obtained by recycling of post-consumer plastic

containers for liquids are very poor. This is due to the incompatibility between component polymers
(PET, PVC, and PE) and degradation of components during the heterogeneous reprocessing.
The use of different classes of additives, such as stabilizers, inert fillers, elastomeric modifiers, and
compatibilizers can improve the processability, enhancing the thermomechanical resistance of the
polymers, and the mechanical properties.
Inert fillers reduce the cost of the recycled plastics but the mechanical properties, although enhanced,
remain insufficient for secondary applications. Elastomers can remarkably improve some mechanical
properties. Some compatibilizing action is contributed by the functionalized polyethylene and sty-
rene-butadiene-styrene rubber and CaO coated with organo-titanates. Finally, good results are ob-
tained for blends produced from this mixture and recycled polyethylene.
INTRODUCTION
The collection of plastic post-consumer containers for liquid is relatively simple
to organize and can be an inexpensive source for recycling of large amounts of
plastic materials. Most plastic containers are made out of three polymers,
namely, high density polyethylene, HDPE, polyvinylchloride, PVC, and
polyethyleneterephthalate, PET.
The presence of these three polymers in the collection of plastic containers
creates difficulty in recycling of such mixed plastic materials. The present tech-
nologies are based on separation of polymers which are then recycled as single
polymers. In general, these recycling plants include different stages: separation
94 Recycling of post-consumer plastic containers for liquids
of HDPE by flotation, identification of PVC and PET containers, automatic sort-

1
ing, washing, and micronization.
The difficulties arising in recycling of mixture of HDPE, PVC, and PET are
due to the incompatibility of these polymers, their different melting points,
o
ranging from 130 to 260 C, and thus different processing temperatures, and
o
degradability of PVC at temperatures above 200 C.
Possible industrial recycling processes of mixed plastics wastes are based
on fast extrusions of these mixtures, having PE as continuous matrix, in ad hoc
2
designed equipment to produce finished products in a single step.
3
A new process is based on the homomicronisation of the polymer mixtures
to obtain a homogeneous new thermoplastic material.
When small amounts of PET are present in the mixture, the processing can
be carried out at temperature below the melting point of this polymer which acts
4,5
as a solid filler. No significant degradation of PVC is observed under these con-
ditions.
6-8
In previous papers, it has been demonstrated that by adding polyethyl-
ene (or recycled polyethylene) to a plastic mixture coming from collection of con-
tainers for liquids, these mixtures with PE content larger than 50% can be
extruded in conventional machineries at high temperatures. The degradation of
PVC is controlled, at last in part, by adding stabilizing agents and by using suit-
able processing conditions.
The mechanical properties strongly depend on composition, and for blends
with PE content larger than 65% are similar to those of the recycled polyethyl-
6,7
ene apart for the elongation at break. Mechanical properties, and in particular
elongation at break and impact strength, can be further improved by adding
8
suitable modifying agents.
Aim of this work is to study the processing in a conventional single screw
extruder of mixed plastics wastes coming from a collection of containers for liq-
uids without any blending with polyethylene and to characterize the resulting
secondary materials. The influence of the processing conditions and the effect of
stabilizing agents, fillers, modifiers, and compatibilizers are also considered.
F. P. La Mantia 95
EXPERIMENTAL
MATERIALS
The post-consumer plastics containers for liquids, PCL, used in this work have
the following composition: PET ≈ 45%, PE ≈ 35%, and PVC ≈ 20%. This composi-
tion is representative of the average composition of the plastic fraction obtained
in 1992 by obligatory separate collection of containers for liquids in Italy. This
composition is different from that obtained in previous years and used in previ-
6-8
ous papers. The difference reflects a lower consumption of PVC for mineral
water bottles replaced by PET, whose concentration in this separate collection is
continuously increasing. The PE content, mainly HDPE, especially used for de-
tergents and cosmetics bottles, is almost constant.
The plastics waste was reduced to small flakes in rotating knives mill.
The thermal stabilizing agents used in the work are reported in Table 1.
Table 1: Thermal stabilizers
Stabilizer Supplier Concentration (%)
Irgafox B603 Ciba-Geigy 0.1
Tribasic lead sulfate - 0.5
Irgafox B603 is a blend of three antioxidants, namely Irganox 1010,

Irganox 1076 and Irgafox 163, suitable for polyethylenes and polyesters.
Tribasic lead sulfate is a typical stabilizer for PVC.
To improve the processing behavior of the mixtures and to reduce the me-
chanical stress in the melt, which can enhance the degradation of the polymers,
two lubricants, see Table 2, were used throughout.
Table 2: Lubricants
Lubricant Supplier Concentration (%)
P520 Hoechst 0.15
Calcium stearate Ciba-Geigy 0.15
Calcium stearate is an external lubricant, while P520 is an apolar low mo-

lecular weight polyethylene wax and then an internal lubricant.
Table 3: Modifiers
Modifier Code Supplier
Kraton G1650 SEBS Shell
Vistalon 3708 EPDM ESSO
Paraloid EXL 3647 MBS Rohm & Haas
FC 45 EVA Enichem
Table 4: Compatibilizers
Compatibilizer and its code Supplier
OCT/AB Micromin
OCT/CH Micromin
SEBS-g-MA Shell
PE-g-MA Enichem
The inert fillers added to the plastics waste were calcium carbonate, saw-
dust and glass fibers. CaCO3 is a powder of about 30 µm in diameter. Sawdust
was obtained by milling of wood in a blade mill through a 20 mesh grid. The glass
fibers had D = 10 µm and length-to-diameter ratio of about 400.
Whereas the fillers have been used especially to reduce the cost of the sec-
ondary materials, the modifiers reported in Table 3 have been used to improve
some mechanical properties.
Kraton G is a SEBS thermoplastic elastomer with a middle block of hydro-
genated polybutadiene and molecular weight of about 64,000. Vistalon is an
EPDM rubber with a molecular weight of about 278,000. Paraloid is a
methylmethacrylate-butadiene-styrene, MBS, elastomer with a core-shell
structure.
FC 45 is a EVA sample with about 14% of EVAc.
Finally, four compounds, see Table 4, have been used in the attempt to en-
hance the adhesion between different phases of the mixtures.
The two OCT samples are small particle size (less than 10 µm) CaO coated
with organo-titanates. In particular AB is coated with neopentyl(diallyl)oxy,
F. P. La Mantia 97
t ri( dioctyl)phos phato ti t a n a t e , w h e re a s CH i s co a t e d w i t h

neopentyl(diallyl)oxy, tri(dioctyl)pyrophosphato titanate.
PE-g-AM is a linear polyethylene functionalized with maleic anhydride.
Melt index is 2.6 and the MA content is about 1.8%.
SEBS-g-MA is a SEBS elastomer functionalized with maleic anhydride
commercially known as Kraton FG 1901.
PROCESSING
All extrusion runs were performed with a laboratory single screw extruder
(D=19 mm, L/D = 25) fitted with a venting port. The blends were mixed by hand
and fed to the extruder. Most of the extrusion runs were done with a tempera-
o
ture profile of 200-230-240-260 C and a screw speed of 80 rpm. Other runs were
carried at a screw speed of 20 rpm.
The extruded materials were cooled in a water bath, granulated and ex-
truded a second time to achieve a good homogenization. As already shown in
6-8
previous works, this procedure gives rise to the best balance between degrada-
tion and homogenization.
MECHANICAL PROPERTIES
The sheets for the stress-strain curves were obtained by compression molding in
o
a laboratory press at 260 C.
The bars for the impact tests were obtained by injection molding in a labo-
ratory molder (Mini Max molder CS 183, Custom Scientific, USA). In both cases
o
the temperature was 260 C.
Stress-strain curves were obtained with an Instron model 1122. Impact
tests were carried out in Izod mode with a CEAST Fractoscope. The results were
averages of at least seven measurements.
PVC DETERMINATION
In order to determine the undegraded PVC content in the recycled material,

o
about 0.2 g of the extruded mixtures were put in a glass tube, heated up to 280 C
o
at a heating rate of about 5 C/min and held at this temperature for about 15 min
under a slight flow of nitrogen. The evolved gases were absorbed in water and
the resulting HCl titrated with NaOH.
The amount of HCl is directly related to the content of undegraded PVC in
the extruded material.
Figure 1. PVC degraded during processing of PLC and PLC with different additives. The dashed
area represents the scatter of the experimental results.
RESULTS AND DISCUSSION

EFFECT OF STABILIZERS AND LUBRICANTS
Figure 1 reports the amount of PVC degraded during two extrusion steps for
PLC and its mixtures with all stabilizing and lubricating additives. When no
stabilizing agent is used, the degradation of PVC is quite extensive: only slightly
more than a half of the initial theoretical PVC content in the raw material is
found in the recycled material. The lubricants do not improve significantly the
stability of PVC which is only slightly enhanced by the antioxidants. This is
somehow surprising since these antioxidants are typical stabilizers for
polyolefines, PE, and not for PVC. In explanation of this result, one must take
9-11
into account the possible interactions occurring during degradation of blends.
The degradation of PVC can be enhanced by the formation of polyethylene radi-
cals, produced by thermomechanical stress but stabilized by these antioxidants.
The slight stabilizing action of the lubricants can be attributed to a lower
mechanical stress acting on the melt during processing. The mechanical stress
can act, indeed, as a “catalyst” of the thermal degradation increasing the level of
the degradation.
An efficient stabilizing action is contributed by the lead compound (which
is a typical stabilizer of PVC) and a mixture of antioxidants and lubricants with
lead salt. With these stabilizing systems, about 80% of the initial undegraded
PVC has been measured in the recycled material.
F. P. La Mantia 99
Figure 2. Elastic modulus of the samples shown in Figure 1.
Figure 3. Tensile strength of the samples shown in Figure 1.
During processing, both PE and PET can also undergo severe degradation
processes. In particular, PE can be degraded by thermomechanical stress and
PET by hydrolytic chain scission since the material is not dried before process-
ing. It is not possible, however, to examine these processes of degradation, and

only the degradation of PVC, the most severe and dangerous, was considered.
Figure 4. Elongation at break of the samples shown in Figure 1.
Figure 5. Impact strength of the samples shown in Figure 1.
The above discussion refers to the extrusion steps performed at a die tem-
o
perature of 260 C and screw speed of 80 rpm. By changing them, for example, in-
creasing temperature and decreasing screw speed, the degradation of PVC
F. P. La Mantia 101
o
becomes more rapid. After extrusion carried out at a die temperature of 260 C
and 20 rpm, small amounts of PVC were detected in the material. Decreasing ro-
tation speed from 80 rpm to 20 rpm, the residence time in the extruder increases
from about 3 min to about 10 min.
The mechanical properties of these materials reflect, at least in part, the
amount of PVC degraded. In Figures 2-5 the elastic modulus, E, the impact
strength, IS, the tensile strength, TS, and the elongation at break, EB, are re-
ported for all samples shown in Figure 1.
Due to a strong incompatibility and severe degradation of the components,
the mechanical properties of PLC are very poor and much lower than those ex-
pected on the basis of an additivity rule for the mixture components.
Table 5: Mechanical properties of the PLC mixture and individual components
Sample E, GPa TS, MPa EB, % IS, J/m
HDPE 5-6 22-24 600-700 750-800
PVC 13-16 6-8 30-40 60-90
PET 12-14 33-36 350-450 40-50
PLC 5-6 4-4.4 1.3-1.6 20-22
In Table 5, the mechanical properties of the individual components (bottle

grade materials) are reported together with those of the recycled mixture. All
mechanical data of PCL are below those evaluated according to the additivity
rule, but elongation at break and impact strength are exceptionally low, indicat-
ing a lack of adhesion between the phases.
Regarding the mechanical properties of the stabilized mixtures, modulus,
impact strength, and elongation at break show improvements up to 25%, but
their values remain low, in particular elongation at break and impact strength.
The above results suggest that an appropriate stabilizing system should be
used to improve properties of the recycled material and in particular to avoid the
degradation of PVC and HCl evolution. HCl can be entrapped in the material,
reducing its resistance and contributing to the pollution of environment.
The next extrusion runs were carried out using the following stabilizing
system: lead stabilizer = 0.5 wt%, B603 = 0.15 wt%.
Figure 6. Elastic modulus of PLC and PLC with 10% inert fillers.
Figure 7. Tensile strength of PLC and PLC with 10% inert fillers.
FILLERS
Inert fillers are used in the thermoplasts to improve some properties of the poly-
mers but, in particular, to reduce the cost of finished products. To estimate both
effects three inert fillers have been tested. Figures 6-9 report the mechanical
F. P. La Mantia 103
properties of the mixture filled with 10% calcium carbonate, glass fibers, and
sawdust. All these fillers give rise to a small increase of modulus and different
effects on the other mechanical characteristics. Glass fibers are very effective in
improvement of all properties, in particular impact and tensile strength are re-
markably enhanced. Significant improvements of the tensile strength are also
observed by adding CaCO3 and sawdust.
Figure 8. Elongation at break of PLC and PLC with 10% inert fillers.
Figure 9. Impact strength of PLC and PLC with 10% inert fillers.
Figure 10. Elastic modulus as a function of fillers content.
Figure 11. Tensile strength as a function of fillers content.
The effect of the concentration on the mechanical properties can be ob-

served from Figures 10-13. All fillers show qualitatively similar results, namely,
the mechanical properties increase up to a content of about 20% then remain
constant or decrease. Concentrations of CaCO3 and sawdust as high as 20% are
economical considering a very low cost of these materials, whereas more expen-
sive glass fibers can be used at a level of 10% to remarkably improve the me-
chanical properties at low cost.
The positive effect on the mechanical properties and on the cost of these fill-
ers is counter-balanced, at least in part, by an increase in the viscosity of the
mixture and then by the increase of the extrusion power or by a reduction of the
F. P. La Mantia 105
output flow rate. The presence of 10% fillers reduces the flow rate of by 25-30%
for all investigated fillers, Figure 14. Increasing the filler content a further re-
duction is observed, Figure 15, but this reduction is very small compared with
that observed at a concentration of 5%. This result suggests that at shear rates
experienced in the extruder the viscosities of the filled melts are very similar, re-
markably larger than that of the unfilled molten mixture and almost independ-
ent of the type and content of filler.
Figure 12. Elongation at break as a function of fillers content.
Figure 13. Impact strength as a function of fillers content.
MODIFYING AGENTS
Better results have been observed by adding 10% of the above described modi-
fier agents, Figures 16-19.
Figure 14. Output flow rate for PLC and PLC with 10% inert fillers.
Figure 15. Output flow rate as a function of the filler content.
The elastic modulus, Figure 16, decreases and this result was expected
from a low modulus of these rubbery compounds. Because of the same reason,
significant enhancement of the other properties is observed. SEBS and EPDM
remarkably improve tensile strength, elongation at break, and impact strength,
Figures 17 to 19, respectively. Elongation at break, in particular, is low and all
the samples show a fragile behavior.
These modifying agents do not act as compatibilizing agents. The improve-
ment of the elongation at break and impact strength of the mixture is mainly
due to the inherent properties of the elastomers. The SEM micrograph of the
system with EPDM, Figure 20a, does not show a better adhesion between the
phases than that observed in non-modified PLC, Figure 20b.
F. P. La Mantia 107
Figure 16. Elastic modulus of PLC and PLC with 10% modifiers.
Figure 17. Tensile strength of PLC and PLC with 10% modifiers.
COMPATIBILIZERS
Some improvement of the mechanical properties is also observed on addition of

compatibilizers, Figures 21-24. The grafted polymers have been used success-
fully as compatibilizers for blends of polyolefines and polar polymers such as
PET and PA-6, and the coated CaO samples gave interesting results in the recy-
12
cling of PET/HDPE blends.
Figure 18. Elongation at break of PLC and PLC with 10% modifiers.
Figure 19. Impact strength of PLC and PLC with 10% modifiers.
Two CaO samples induce a significant improvement of modulus and tensile

strength, but only a modest improvement of impact strength. The two grafted
samples, and in particular the SEBS rubber, are particularly effective for im-
proving tensile and impact strength. The mechanical properties are enhanced
by adding these materials and these improvements can be, in this case, attrib-
uted to a better adhesion between the three phases. SEM micrographs of PLC
with OCT/CH and grafted PE, Figures 25 a and b, show, indeed, a more homoge-
neous morphology compared with that of the non-compatibilized PLC, Figure
20b. In particular, a minor number of voids is observed, and, moreover, the di-
mensions of the dispersed particles are slightly lower.
F. P. La Mantia 109
Figure 20. SEM micrographs: (a) PLC with 10% EPDM, (b) PLC.
COMPARISON BETWEEN DIFFERENT SYSTEMS
In this paragraph, the results of the above discussed systems are compared with
those for a blend obtained by adding 50% of recycled polyethylene, RPE, to the
PLC mixture. RPE is mainly composed of low density polyethylene coming from
recycling of greenhouse films. This new system has been investigated consider-
6,7
ing the results already obtained for similar compositions.
Figure 21. Elastic modulus of PLC and PLC with 5% compatibilizers.
Figure 22. Tensile strength of PLC and PLC with 5% compatibilizers.
Other systems were also selected, considering mechanical properties,

processability, and cost of additives. In particular, CaCO3 is an inexpensive
filler, PE-g-MA and OCT/CH are also relatively inexpensive, and give some im-
provement of the mechanical properties. EPDM is quite expensive but it contrib-
utes to a remarkable enhancement of elongation at break and impact strength.
F. P. La Mantia 111
Figure 23. Elongation at break of PLC and PLC with 5% compatibilizers.
Figure 24. Impact strength of PLC and PLC with 5% compatibilizers.

Figure 25. SEM micrographs: (a) PLC with 5% OCT/CH and (b) PLC with 5% PE-g-MA.
The processability, as evaluated through the output flow rate, Figure 26,
decreases with filler addition, but only by adding 20% of CaCO3 the output flow
rate is rapidly reduced. All other additives and the high viscosity RPE cause
only to a small decrease of the processability of the PLC compound.
As for the elastic modulus, Figure 27, only the coated CaO improves modu-
lus; the decrease observed for the blend with RPE is due to a low modulus of this
recycled material as compared with PLC.
F. P. La Mantia 113
Figure 26. Output flow rate of different systems.
Figure 27. Elastic modulus of different systems.
All additives and RPE improve the tensile strength, Figure 28, whereas
only RPE increases elongation at break, Figure 29.
EPDM and RPE offer substantial enhancement of impact strength,
Figure 30.
Figure 28. Tensile strength of different systems.
Figure 29. Elongation at break of different systems.
CONCLUSIONS
• The recycling of complex, heterogeneous mixtures of polymers is an impor-
tant challenge for the recycling of post-consumer plastics. Incompatibility
and degradation render secondary materials with poor mechanical proper-
ties.
F. P. La Mantia 115
Figure 30. Impact strength of different systems.
• The results shown in this work indicate that three components blends (PE,
PET, PVC) can be reprocessed several times at high temperatures and the
degradation of PVC can be controlled by using suitable stabilizing systems
and processing conditions. The mechanical properties, however, are not
sufficient. Some improvements can be achieved, using additives such as in-
ert fillers, elastomers, and compatibilizers. Calcium carbonate and saw-
dust lower the cost of the recycled materials but induce only limited
enhancement of some mechanical properties. Addition of glass fibers al-
lows to improve tensile and impact strength.
• Elastomers, used as modifiers (EPDM and SEBS), are very effective in im-
proving the ultimate properties of the secondary materials. These improve-
ments are mostly due to inherent characteristics of the elastomers which
act as a forth phase in the blend.
• The tests carried out with some functionalized compounds suggest that
maleic anhydride grafted polymers and CaO coated with organo-titanates
act as compatibilizers modifying the morphology of the blend but the im-
provement of the mechanical properties does not fully justify cost of
compatibilization.
• The addition of considerable amount of recycled polyethylene to the mix-
ture results in a blend, having polyethylenic matrix, which improves me-
chanical properties.
ACKNOWLEDGMENT
This work has been financially supported by MURST (60%) .
REFERENCES
1. E. Sereni in Plastic Materials Recycling, Ed. F. P. La Mantia, ChemTec Publishing,
Toronto, 1993.
2. J. Leidner in Plastics Waste, M. Dekker, New York, 1981.
3. B. Dubrulle d’Orhcel, ReC’93, 17-20 January, 1993.
4. F. Cremonini, L. Castelnuovo, European Regional Meeting of Polymer Processing
Society, Palermo (Italy), September 15-18, 1991.
5. A. Vezzoli, M. Lamperti, and C. A. Beretta in Plastic Materials Recycling,
Ed. F. P. La Mantia, ChemTec Publishing, Toronto, 1993.
6. F. P. La Mantia, Polym. Deg. Stab., 37, 145 (1992).
7. F. P. La Mantia, C. Perrone, and E. Bellio in Plastic Materials Recycling,
9. I. McNeill in Developments in Polymer Degradation, Ed. N. Grassie, Applied
Science Publishers, London, 1977.
10. F. P. La Mantia in Handbook of Polymer Degradation, Eds. S. H. Hamid,
M. B. Amin, and A. G. Maadhah, M. Dekker, New York, 1992.
11. F. P. La Mantia, International Conference on Advances in the Stabilization and
Controlled Degradation of Polymers, Luzern, 25-27 May, 1988.
12. F. P. La Mantia and M. Vinci, Polymer Recycling, 1 (1994).
B. Dubrulle D'Orhcel 117
A New Process for Recycling of

Mixed Plastics Waste
B. Dubrulle D’Orhcel
Process Newplast Holding N.V., Herengracht 320, NL-1016 CE Amsterdam, Holland
A new process for recycling mixed plastics waste is described. With this process mixed plastics waste
can be recycled without any compatibilizer giving rise to secondary materials easily processable in
conventional machinery and with good mechanical properties. The process is based on
homomicronization, a process of deep microhomogenization which permits differing polymers to re-
sult in a solid amalgam with good physical properties.
INTRODUCTION
Recycling of mixed plastics waste is very difficult, in particular because of
the incompatibility of different polymers and danger of degradation of polymers,
1,2
such as PVC, but also due to their different melting points.
Some industrial or laboratory attempts are based on reprocessing at low
temperatures. In these attempts high melting polymers are used as solid fill-
3
ers. Another useful method comprises the use of compatibilizers and stabilizers
4-8
in the polymer blends. Mechanical properties of these recycled mixed plastics
are generally inferior, the surface properties are poor, and in some cases the cost
of compatibilizers and stabilizers is too high for recycled materials.
The possibility of recycling of mixed plastics waste without any
compatibilizer and obtaining a recycled material which easily processed in con-
ventional machinery add value to manufacture. The properties of resultant ma-
terial are also essential.
118 A new process for recycling of mixed plastic waste
Following is a description of a new process aiming at achieving this goal.

The pr oc es s is b a s ed on h o m o m i cro n i z a t i o n , a pro ce ss o f de e p
microhomogenization which permits differing polymers to result in a solid
amalgam with good physical properties without prior sorting of the different
polymers and with a very attractive cost.
THE NEWPLAST-PROCESS TECHNOLOGY

The NEWPLAST-PROCESS line consists of three modular zones. Each
zone can be entirely automated to produce a compound made from mixed plas-
tics waste from industrial and/or urban sources. In the first area (Zone I) (Figure
1), waste is prepared for processing. The equipment used depends on the con-
tamination level of the feedstock materials.
In the first processing step, the material is shredded. Ferrous and non-fer-
rous materials and heavy elements are separated. This is followed by grinding,
washing, paper separation, and drying. The prepared mixture is now clean, dry
and of consistent size.
This material is conveyed to homogenization silos and its composition is
validated by infrared spectrometry. If necessary, the mixture can be supple-
mented with materials which are at too low concentration than expected and/or
additives for a specific application.
The resulting mixture is then pneumatically transported into the second
area (Zone II) (Figure 2) and automatically transferred into the
homomicronization chamber.
Intense mechanical shearing causes a rise in temperature (about
o
200-230 C depending on the polymer composition) and the molecular linking of
various polymer chains. This produces a homogeneous dough. The dough is ex-
truded to filtering and degasification and then processed into pellets. The re-
TM
sulting Newplast compound has chemical and mechanical properties that are
comparable to certain virgin resins. The compound can be processed with stan-
dard plastics techniques.
A third area (Zone III) (Figure 3) receives the pellets for conditioning. The
TM
Newplast compound is automatically transferred to the integrated produc-
tion equipment for continuous parts manufacture. The energy requirements of
the NEWPLAST-PROCESS are comparable to those of virgin-resin installation
TM
of identical capacity. The various life cycles of the Newplast compound give a
positive balance.
Figure 1. Scheme of the Zone I - reception and pretreatment of contaminated plastics waste. Plan
of Newplast-Process compounding line of heterogeneous plastic waste - capacity 1000 kg/h.
Figure 2. Scheme of the Zone II - homogenization.
The NEWPLAST-PROCESS technology can be easily adapted to different

plastic mixtures or and/or formulated for a specific application. Such flexibility
is made possible because the elements required for the final application can be
added during processing.
DESCRIPTION OF A NEWPLAST-PROCESS INSTALLATION

The Newplast recycling line is divided into several sections, each fulfilling
a particular function within the process. It has been designed to manage to treat
waste containing up to 30% of contamination by paper, wood, sand, and metallic
parts.
Zone I. Reception and pretreatment of contaminated plastics waste
The main items of this zone are (see Figure 1):
• Shredder
• Magnetic separator (ferrous metals)
• High density foreign parts separator (glass, stone)
• Intensive washing
Figure 3. Scheme of the Zone III - packaging.
• Grinder
• Friction separator (paper, cardboard)
• Intensive washing
• Mechanical dryer
• Thermal dryer
• Non-ferrous separator
• Grinder
Zone II - Homomicronization Zone
The main items of this zone are (see Figure 2):
• Dosing/weighting
• Homomicronizer
• Filtration and degassing extruder
• Pelletization
Zone III - Packaging Zone
• The main items of this zone are (see Figure 3):
• Storage silos
• Bagging unit
• Palletizing unit
HOMOMICRONIZATION EQUIPMENT
The homomicronizer is the heart of the line. It transforms plastics waste
into Newplast. It includes all means necessary to guarantee the safest possible
operation within the context of the latest industrial safety criteria.
It consists of a cylindrical chamber, made from a special steel, within which
a rotor turns at high speed. This has specially designed, angled blades which in-
timately micronizes and homogenizes the mixed thermoplastics particles. The
rotor is electrically driven by a computer-controlled, variable-speed motor pro-
grammed adequately to materials processed.
The device shown, in the Figure 4, comprises a cylindrical chamber [2] with
a horizontal axis, provided with a feeding hopper [4] placed in an upper position
and a discharge well [5]. A rotor [6], mounted co-axially inside the chamber [2],
is operated by means of an electric motor [7]. The walls of the chamber [2] consist
of an assembly [10] formed by uniting two semi-cylindrical shells [11 and 12], as-
sembled together by means of a hinge along one lateral edge [13] and a safety
closing device [14] is disposed along the edge [15] opposite the edge [14].
Figure 4. HMC device.

Figure 5. Rotor of the HMC device.
As can best be seen in Figure 5, the rotor [6] comprises a cylindrical shaft
[30] provided with a plurality of radial blades [31, 32], the shapes of which are of
two different types. More precisely, the orientation of the blades [32], mounted
in the vicinity of each end of the shaft [30], is such that it acts to impart to a load
of material being treated in the chamber [2] a motion which tends to push this
material away from the chamber wall adjoining the ends of the shaft [30] to
bring it back towards the inside of the chamber, through the region subjected to
the agitation and stirring action of the blades [31].
The diameter of the chamber [2] and the length of the blades [31 and 32] are
such that, when the rotor [6] is at room temperature or in the vicinity of this tem-
perature, the clearance between the tips of the blades [31 and 32] and the inner
wall [40] of chamber [2] is in the order of 0,5 to 1 mm.
It has been found advantageous that the rotor [6] and its drive motor [7] be
dimensioned relatively to the inner volume of the chamber [2], so that the maxi-
mum mechanical power transferable to the material being treated in the cham-
ber [2] by the rotor [6] be in the order of 1 to 2 kW for each liter of this material.
For example, for a total inner volume of the chamber [2] of 85 liters (case of an
experimental prototype), it was found that the maximum power delivered by the
motor is in the order of 128 kW.
The mixed plastics particles to be treated are fed to the hopper located on
the top of the machine. This releases the material into the drum via a rotary,
pneumatically-operated cover made from a special material. The cover’s two
pneumatic actuators are locked in the closed position, during processing, by a
security lock. The cover’s seal is specifically manufactured to prevent any leak-
age. The mixed plastics are melted by the friction developed between the parti-
cles as well as that added by the rotor blades and chamber wall. The molten state
is sensed by the temperature probes and by a sharp increase of the engine
torque. At this time the molten mixture is completely homogenized and is dis-
charged from the chamber.
NEWPLAST-PROCESS
The process and the machine are computer controlled with a specific pro-
gram which has been developed for each application. Parameters determining
each phase of the process are:
• Temperature
• Rotor speed associated with torque and torque variation
• Time
During the process, the computer receives the above information and man-
ages the rotor speed, the rate of the acceleration and deceleration of the speed
and reads the temperature of the rotor and chamber’s wall.
Depending on the type of mixture and the capacity of the machine, the cycle
can take between 35 and 120 s. Once the gelification phase is completed, the un-
loading door opens, combined with a decrease of the rotation speed of the rotor,
and the mixture is then transported by feeding force onto extruder screw. At this
stage of the process, according to the type of the mixed polymers, the tempera-
o
ture of the mixture varies between 200 and 230 C.
The homomicronizer is fed by an automatic batch delivery device, which, in
turn, is fed from a silo of mixed plastics wastes, shredded, washed, dried, and
stocked. The conditions of these wastes are (for a given production cost):
organic impurities (butter, milk, shampoo) 0%

moisture maximum 1%
non-thermoplastics maximum 3%
uniform pellets size.
Figure 6. Mechanical properties of NPP and virgin materials.
CHARACTERIZATION OF THE MIXED RECYCLED POLYMERS

Mechanical properties of the mixed recycled polymers are dependent on
the feed composition. This means that it is possible to produce a large variety of
secondary materials with a wide range of mechanical properties. We will report
the characterization of two different compounds obtained by the above de-
scribed process beginning with different feed composition, reported in Tables 1
and 2. 2.8% of a plasticizer, DOP, and 0.6% of an antioxidant, Irganox 1010, have
been added to both-materials.
The main mechanical properties, obtained according to ISO R257, are re-
ported in Figure 6. In the same Figure the mechanical properties of virgin high
impact polystyrene and virgin HDPE samples are also reported for comparison.
The first virgin material has been compared with the sample NPP28 VO which
shows a fragile behavior, whereas HDPE has been compared with NPP40 AO
which, on the contrary, is a ductile material, see the stress strain curves of these
samples in Figure 7.
Table 1: Composition of mixture NPP28 VO
Component Concentration, %
HDPE 63.75
PVC 8.75
PP 8.75
PS 8.75
PET 5
ABS 5
Table 2: Composition of mixture NPP40 AO
Component Concentration, %
HDPE 63.75
PP 11.25
PVC 4.5
ABS 4.5
PS 4.5
PET 4.5
PP 5.0
EVA 2.0
The modulus, E, of the recycled mixed plastics compounds is higher than

that of virgin HDPE because of the presence of high-modulus components (PET,
PVC) in the mixtures. Both tensile strength, TS, and yield stress, YS, are very
similar than those of the virgin materials. The elongation at break, EB, of the
two recycled samples are lower than those of the relative virgin polymer re-
ported for comparison. It is clear that it is possible to produce secondary materi-
als, such as NPP 40 AO, from incompatible blends with acceptable values of the
elongation at break. This comparison could be even more favorable if experi-
ment was done with the recycling of homopolymers rather than processing ac-
tual wastes.
Figure 7. Typical stress-strain curves of NPP and virgin materials.
CONCLUSIONS
The main advantage of this technology is that it can process heterogeneous
mixed plastics wastes from urban waste into a thermoplastics compound with
well accepted physical properties. The product can be obtained with a total pro-
cessing cost which is today 30-40% of a virgin polymer cost of similar properties.
Due to industrialization of the process, the only sure feedstock source of
large quantities of plastics wastes is from urban wastes, not only in large quan-
tity but also with a certain consistency in composition.
Finally, this thermoplastics compound can be processed, once formulated,
by all the standard plastics processing equipments.
REFERENCES
1. J. Leidner, Plastics Waste, M. Dekker, New York, 1981.
2. F. P. La Mantia, this book, p. 63.
3. A. Vezzoli, M. Lamperti, and C.A. Beretta in Recycling of Plastic Materials,
5. F. P. La Mantia, C. Perrone, and E. Bellio in Recycling of Plastic Materials,

7. B. Ballauri, M. Trabuio, and F. P. La Mantia, this book, p. 77.
8. F. P. La Mantia, this book, p. 93.
B. Boutevin, Y. Pietrasanta, and J.-J. Robin 129
Grafted Copolymers and Their Use

in Polymer Blends
B. Boutevin, Y. Pietrasanta, and Jean-Jacques Robin

Ecole Nationale Superieure de Chimie de Montpellier,
URA - CNRS 11930, 8, Rue Ecole Normale, 34053 Montpellier Cedex, France
The present study summarizes the main radical methods of synthesis of grafted copolymers. Among
these techniques, ozonization is an original and simple way to obtain this type of copolymer. The
ozonization procedure is described in details together with the methods used to characterize activated
polymers, species obtained, and grafting techniques. The study also presents the main applications of
the grafted copolymers obtained by this technique. Various mixtures of incompatible polymers can be
compatibilized by these copolymers which clearly enhances the global performances of the materials.
The crosslinked polymers have been decrosslinked by ozone, which allowed their evaluation, espe-
cially in compositions with bitumens. The ozonized polyethylene wastes can be used in bitumens for
reinforcement. In all the examples dealt with, the study describes the analysis methods and the prop-
erties of the obtained materials and shows that ozonization constitutes an original way of treatment
and utilization of plastic wastes.
1 INTRODUCTION
Over the last years, plastic recycling has become a major concern for industrial-
ists and public because of the increasing consumption and the decrease in avail-
able landfilling areas. It was necessary to find alternative routes which can help
to retain the energetic content of these materials but also preserve our environ-
ment. During last years, many research works have been funded in this field.
The recycling operation is now an attractive concept but not always very easy to
accomplish, especially in the following cases:
• mixtures of polymers difficult to separate for technical or economic rea-
sons,
• combinations of polymers obtained by coextrusion and other methods
• crosslinked polymers.
130 Grafted copolymers and their use in polymer blends
These well defined cases require special treatment. In the case of mixtures,
compatibilization is a feasible solution because it modifies polymers which are
not naturally miscible into compatible or partially compatible polymers. The use
of compatibilizing or emulsifying agents is important because they allow to re-
duce the phases size and ensure their anchoring. Moreover, this technique is
broadly used to obtain mixtures of virgin polymers called alloys. In the case of
crosslinked polymers, the works to use these products are conducted in several
directions:
• decrosslinking and depolymerization of the polymers to obtain basic mono-
mers
• grinding of materials and incorporation of the ground species as fillers in
formulations.
The decrosslinking alone is difficult to achieve and there are a few refer-
ences regarding this method. In this chapter, a method, optimized by our team
to utilize crosslinked polymers, is presented.
A great part of the research carried out by international teams is centered
on the compatibilization of polymers. Number of techniques have been devel-
oped depending on the polymer composition. They may be summarized as fol-
lows:
• addition of a compatibilizer to polymer blends before processing. The
compatibilizers used are, in most cases, block or grafted copolymers
• addition of a compatibilizer during processing. In this case, the additives
are mostly copolymers or polymers having functional groups such as
maleic anhydride, or simply smaller reactive molecules (e.g., maleic anhy-
dride) added in the presence of an initiator, or functional mineral fillers
able to play a role of a compatibilizer.
In this chapter, we are dealing with agents added before processing, and
more particularly with grafted copolymers. The original method to synthesize
this type of copolymers was developed and applied to polymer blends. The
method to obtain these copolymers and their application to various binary mix-
tures of polymers is discussed below. The synthesis method uses the ozonization
reaction for decrosslinking of radical crosslinked polymers and preparation of
polymers reactive towards bitumens. These last works have been conducted to-
gether with Professor M. Delmas and his co-workers from Ecole de Chimie at
Toulouse.
2 THE MAJOR METHODS OF SYNTHESIS OF GRAFTED COPOLYMERS

Grafted copolymers and block copolymers have in the same macromolecule
monomers of different nature and properties. They enable one to obtain the ma-
terials having special characteristics such as phase segregation. Furthermore,
this particular class of polymers can be synthesized by several routes which ren-
der them very attractive. Two main routes may be considered to prepare the co-
polymers: the ionic route and the radical route. To be complete on this subject, a
third route via addition or condensation reactions can also be mentioned. We
have selected these last two reactions because they are easy to perform. In a first
part, we will describe various techniques which can be considered in these two
categories, before more detail is given to one of these methods, namely,
ozonization.
Grafted copolymers can be prepared by different methods, such as:
• combination
• macroradicals
• transfer reaction to the polymer
• macroinitiators
• macromonomers
2.1 THE COMBINATION REACTION
This reaction may be performed in bulk or in a molten state for two polymers,
having reactive groups in their structure. If a polymer polymer A includes func-
tional groups X, distributed along its chain, and a polymer B includes groups Y
at its ends, it is possible to obtain a grafted copolymer by reaction between X and
Y:
It is a typical reaction. For example, the molten mixture of a styrene-maleic

anhydride copolymer and a copolymer of polyethylene grafted by a tertiary or a
1,2
secondary amine leads to the formation of a P(EgS) grafted copolymer.
3
Lambla et al. grafted monohydroxyl oligomers of polystyrene onto copolymers
4
based on styrene and maleic anhydride, using this method. Parsy et al. ob-
tained grafted copolymers by reacting amino-polyamide oligomers with
polyolefins grafted with succinimic groups. This type of reaction is used to
compatibilize immiscible polymers via the in situ formation in grafted copoly-
5
mer blends.
2.2 MACRORADICALS
The radical polymerization of a monomer A initiated by a macroradical on the

side of a polymer chain, leads to the formation of a grafted copolymer according
to the following reaction scheme:
The recombination of two macroradicals having different natures leads to a

grafted copolymer. These macroradicals can be obtained through several meth-
ods, discussed below.
2.2.1 Degradation of polymers
2.2.1.1 Mechanical degradation
The simultaneous degradation of several vinyl or diene polymers produces radi-
cals capable of reacting with each other by a cross-combination, leading to the in
situ formation of grafted copolymers. This degradation may be carried out by
6
high temperature mastication or polymerization of monomer in the presence of
7
polymer and peroxide-type initiator. Using such a method, it was possible to
6
prepare polyethylene and polyisobutene grafted copolymers and acrylic and vi-
7
nyl based monomers.
2.2.1.2 Ultrasonic degradation

Polymers can be degraded by ultra-sounds. The ultra-sound treatment of a solu-
tion of methyl methacrylate and styrene homopolymers gives a certain amount
8
of grafted copolymer.
2.2.2 Chemical route
Some macroradicals are obtained by chemical reactions such as oxidation-re-
duction or diazotation.
2.2.2.1 Oxidation-reduction
Transition metals, such as chromium, manganese, or cerium are able to oxidize
organic functional group, such as alcohols, aldehydes, amines, thiols, forming
9
free radicals which can initiate copolymerization:
Ce 4+ + RCH 2OH → RC • HOH or RCH 2O • + Ce 3 + + H +

10
By this method, methyl methacrylate have been grafted onto PET or
11
acrylonitrile onto polyvinyl alcohol.
2.2.2.2 Diazotation
It is possible to obtain a macroradical from a polymer bearing aromatic rings
12
(e.g., polystyrene) by nitration followed by diazotation. Cooper proposed the
following reaction mechanism:
RC 6 H 4 NO 2 → RC 6 H 4 N 2+
RC 6 H 4 N 2+ + Fe 2+ → RC 6 H 4 N •2 + Fe 3 +
RC 6 H 4 N 2• → RC 6 H 4• + N 2
13 14
The method was also used by Valentine and Hanh to prepare poly(sty-
rene-g-acrylonitrile) copolymers.
2.2.3 Irradiation
In many cases, the chemical modification of polymers by UV, γ-radiation, or elec-
tron bombardment constitutes a way to form macroradicals, allowing the syn-
thesis of grafted copolymers. Number of studies refer to macromolecular
15,16
chemistry under γ-rays or accelerated electrons irradiation. These reactions
are carried out in the absence of oxygen to avoid formation of peroxides or
hydroperoxides. Various polymers can be activated by this method, including
polyolefins, and also more resistant polymers such as PVDF or PTFE. For exam-
17 18,19,20,21
ple, through this technique, acrylic acid or styrene may be grafted onto
22
polyethylene or chloromethyl styrene can be grafted onto teflon and vinyl
23
pyridine onto PVDF.
2.3 TRANSFER TO POLYMER
If a monomer B is polymerized in the presence of a polymer A, already formed,

transfer reactions may occur with respect to the polymer A chain, leading to
24
grafted copolymers. This type of reaction was studied by Shultz, who deter-
mined, for the first time, the transfer constant to the polymer. Other copolymers
25,26
were also prepared by this method, e.g., vinyl chloride and acrylic and
27
methacrylic monomers were grafted onto polyethylene.
2.4 MACROINITIATORS
Macroinitiators are polymers bearing groups able to decompose under the ac-
tion of heat or light to produce free radicals on the chain. Peroxides and
hydroperoxides are the best known groups which, after decomposition of the ox-
ygenated groups, initiate the polymerization of any monomer present in the me-
dium and are polymerizable by radical route. Macroinitiators can be prepared
by different methods discussed below.
2.4.1 Oxidation
This technique was developed, for the first time, at Brooklyn Polytechnic Insti-
28
tute, where hydroperoxides, resulting from the thermal oxidation of a vinyl
polymer, were used to initiate the polymerization of vinyl monomers. The graft-
ing of styrene, methyl methacrylate, or vinyl chloride onto poly-α-olefins was re-
29
alized by this method. The thermal oxidation mechanism was shown to involve
the formation of hydroperoxides, which decompose to a number of oxygenated
30
functional groups, not necessarily leading to a grafted copolymer.
2.4.2 Ozonization
Contrary to oxidation, ozonization forms relatively stable peroxide and
hydroperoxide groups which can be used to initiate a further polymerization.
Many grafted copolymers were synthesized by this technique, especially
31 32
polyolefins grafted by fluorinated monomers or acrylic monomers, polyvinyl
33
chloride grafted by styrene, and fluorinated polymers grafted by styrene or
34,35
acrylic monomers.
2.4.3 Preirradiation in the presence of air
In this case, the polymer is exposed to an ionizing radiation, γ −rays, generally,
in the presence of oxygen or air. The radicals, which are formed, react with oxy-
gen to give hydroperoxides and peroxides which, by heating in the presence of
monomers, lead to the formation of grafted copolymers and homopolymers.
36
Chapiro showed that room temperature irradiation of polyethylene gives per-
oxides whereas polypropylene irradiation results in hydroperoxides under the
same conditions. By this technique, it was possible to graft styrene, methyl
36
methacrylate, and acrylonitrile onto polymers, such as polyethylene, poly-
37 38
propylene, and polyvinylchloride.
2.5 MACROMONOMERS
Macromolecular monomers or macromonomers are oligomers or polymers hav-

ing a polymerizable functional group at the end of chain. The double bond allows
copolymerization reactions in which the length and the number of grafts ob-
tained can be controlled. Macromonomers can be synthesized by ionic route but
also by radical route as discussed below.
2.5.1 Radical polymerization
The use of transfer agents in polymerization reactions permits one to obtain
macromonomers. The following scheme shows the synthesis of MMA
39,40
macromonomer by using thioglycolic acid as a transfer agent, and reaction
with glycidyl methacrylate:
The major part of the work was accomplished, using mercaptan-type trans-
41-45
fer agents with acrylic or methacrylic monomers, or monomers such as vinyl
46,47
chloride.
2.5.2 Polyaddition reaction
Polyaddition of non-conjugated dienes and diamines, under stoichiometric con-
ditions, leads to the formation of polyamine macromonomers. For example,
polyaddition of divinylbenzene and a secondary diamine gives a polyamine
48,49
macromonomer:
The above review shows that there is a variety of methods to prepare

grafted polymers. We have selected, for our work, the method using
macroinitiators prepared by ozonization of polymers.
3 SYNTHESIS OF GRAFTED COPOLYMERS BY OZONIZATION OF POLYMERS
3.1 MECHANISM OF OZONIZATION
The reaction of ozone with organic compounds has often been studied. It is an in-
teresting way to introduce oxygenated compounds into structures, and, conse-
quently, to form C-O-C bonds. The reaction produces several compounds,
difficult to separate, which is the main drawback of this reaction. Many parame-
ters are involved in this type of oxidation:
• reaction environment (solution, gaseous phase, fluidized bed for solids)

• nature of solvent
• reaction temperature
• exact composition of the O2/O3 mixture
• reaction time.
Our purpose in these studies was aimed at the modification of polymers, in
order to make them reactive towards monomers (eventually, towards reagents
such as bitumen), rather than the preparation of products of very high purity.
Generally, the ozonization reaction of polymers or saturated organic com-
pounds proceeds as follows (different mechanisms are proposed by various au-
thors who took interest in the subject):
RH + O 3 → RO • + HO 2• (50)
RH + O 3 → R • + O 2 + HO • (51)
RH + O 3 → ROOOH → RO • + HO 2• (52)
also O 3 → O 2 + •O •
and RH + •O • → R • + HO • (53)
Another hypothesis formulated in a study on ozonized polyethylene (based

54
on ESR results) leads to the following scheme:
The above processes constitute the first step of the reaction. They result in
rearrangement or decomposition secondary reactions forming acid, ketone, or
55-59
aldehyde compounds. The ozonization reaction of double bonds is much
faster and gives rise to ozonides which also decompose and yield many second-
61-63
ary products. In the case of aromatic rings instable triozonides are obtained
64,65
and they decompose very quickly.
The reaction may be carried out for dissolved polymers by bubbling O2/O3
gaseous mixture in appropriate solvents (solvents which are not decomposed by
ozone) or in bulk, using finely ground polymers suspended in a fluidized bed.
Polymers should be ground to a very fine powder because the peroxidation reac-
tions occur at the grains periphery and are substantially retarded in the grain
core due to limits of gas diffusion. The crystallinity of the basic polymer is of par-
66
ticular importance because it decreases diffusion rate in the crystalline zones.
Generally, the reaction is carried out at room temperature, but it may also be
o
conducted at moderate temperatures (<100 C).
For the most part, the ozone generators deliver gaseous mixtures contain-
ing more or less ozone depending whether ambient air or pure oxygen are used
as feed gas. The ozone flow is generally in the range of 18 to 36 g/hr.
3.2 DETERMINATION OF THE POLYMER ACTIVATION
The main functional groups of interest in this study, formed during the
ozonization, are peroxides and hydroperoxides and their formation must be
monitored.
3.2.1 Colorimetric titration
67
This method, described by T. Sarraf, involves a free radical scavenger: diphen-
yl picryl hydrazil, DPPH. DPPH, stable at room temperature, is added in excess
o
to the ozonized sample and reacted at 110 C. The cleavage of peroxides and
hydroperoxides to RO∴φ9 and HO∴φ9 generates free radicals which are captured
by DPPH. After precipitation of polyethylene, the excess of DPPH in the filtrate
is back-titrated by colorimetry at 520 nm, and the amount of initial peroxides
and hydroperoxides can be obtained from the result.
3.2.2 Selective titration of hydroperoxides in the presence of ferrous salts
68
The method was introduced by Landler and Lebel who determined the decom-
position constants of ozonized polyethylene, polyvinylchloride, and polystyrene,
69
without distinction between peroxides and hydroperoxides. Zeppenfeld dem-
onstrated that only hydroperoxides in acidic medium can oxidize ferrous ions ac-
cording to the following reactions:
ROOH + Fe 2+ + H + → RO ± + Fe 3 + + H 2O
RO • + Fe 2+ + H + → ROH + Fe 3 +
2+
It should be noted that a hydroperoxide molecule consumes 2Fe . The ti-
3+
tration is done after complexation of the ferric iron (Fe ) by ammonium
thiocyanate, leading to a red complex readily titrated by colorimetry at 512 nm.
3.2.3 Iodometric titration
70-72
Even if it is not universal, it is the most mentioned method. It is based on oxi-
dation by the peroxides in acidic medium of the iodide ions into iodine. The reac-
tion scheme is as follows:
ROOR′ + 2I- + 2H + → ROH + R′ OH + I 2
The iodine generated in the solution is titrated by sodium thiosulfate in the

presence of starch-paste
I 2 + 2S 2O 32 − → 2I- + S 4O 26 −
This method is simple but it has many constraints, the main is the need for
the very strict experimental conditions. Effectively, a side-reaction may occur:
addition of iodine to the double bonds of the titrated polymer (either formed dur-
ing the ozonization or present in the starting polymer).
3.3 DETERMINATION OF THE DISSOCIATION CONSTANT OF PEROXY FUNCTIONS
Species formed during the ozonization behave like classical initiators. It was,
thus, interesting to evaluate their decomposition rate constant Kd. For this pur-
pose, several methods were used by various authors working on this subject.
3.3.1 Determination of the global dissociation constant Kd
The peroxide functional groups decompose into free radicals according to the re-
action:
ROOH → 2RO •
ROOH → RO • + HO •
It is necessary to know the amount of species when grafted copolymers are

67 34
prepared. Sarraf et al. and Serdani et al. have set up a method which allows
one to follow the amount of free radicals formed as a function of time. The
amount of peroxides and hydroperoxides, Nt, in the PE and PVDF samples, kept
at a given temperature, was titrated by colorimetry using DPPH as a function of
time. Without distinction between the two species, they deduced the amounts of
non-decomposed peroxides or hydroperoxides as a function of time. Assuming
that the decomposition of these species follows the classical laws of decomposi-
tion of initiators, I 2 = [I 2 ]o e -k d t , the plot of Ln No/Nt = f(t) should give a straight
line, and its slope the value of the decomposition constant, Kd. Moreover, from
the calculation of Kd at several temperatures, the value of the activation energy
1/T
was obtained by plotting LnKd = f . This method may be simple, but it has the
disadvantage of neglecting the real efficiency factor, f, for peroxidic species.
Also, the decomposition constant of these functional groups is measured in a
special medium, in the presence of a stable free radical (DPPH), which is very
different from the medium encountered during grafting reactions.
3.3.2 Determination of Kd constant and the efficiency factor, f
The decomposition of the free radicals produced from peroxide or hydroperoxide
functionality, according to the reaction scheme previously described, does not
necessary lead to the initiation of a grafting reaction. The efficiency of each radi-
cal, to generate a new macromolecular chain, must be accounted for. The initia-
tion rate can be written as follows :
Va = 2fKd [I2] with f the efficiency factor of the initiation

73
This term was neglected by previous authors and recently, Fargère et al.
have set up a kinetic method allowing direct calculation of the real Kd value and
f. For this purpose, the authors conduct the grafting of ethylene-vinyl acetate co-
polymer by styrene (the latter plays role of the solvent). Following the changes
in styrene concentration in the reaction medium as a function of time and apply-
ing the kinetic laws of radical polymerization, they deduce that:
2 1/2 1/2 1
at t = 0 (d(Ln[M]o/([M])/dt) = (kp /kte) (fKd) [I2]o2
Plotting Ln[M]o/[M] = f(t) for low conversion rates, for each grafting temper-
ature, a curve is obtained from which the slope of the origin tangent allows the
calculation of fKd. The authors show that the influence of the monomer concen-
tration at the beginning is not negligible, and it determines the reaction medium
viscosity. Moreover, the grafting reactions are accompanied by the thermal
homopolymerization reactions which have been taken into account by the au-
thors in evaluation of their experiment and contribution to the monomer disap-
pearance rate.
The real activation energies of the grafting reaction, initiated by peroxides,
was calculated. The efficiency factor was calculated from a low conversion rates
of monomer by comparing the theoretical curve obtained from the general ex-
pression of the monomer disappearance in function of time to the real curve.
This comparison allowed authors to evaluate f and Kd.
3.3.3 Comparison between the dissociation constants

The values of the dissociation constants obtained by all the authors are rela-
tively high for radical initiators. They are as follows :
75 3
Ozonized PE Kd = 42.4×10 exp(-6300/T) and Ea = 52.5 KJ/mol
76
Ozonized PVDF Kd = 338exp(-4700/T) and Ea = 39 KJ/mol
77
Ozonized EVA Kd = 1481.5exp(-5500/T) and Ea = 46 KJ/mol
Knowing that for classical initiators, the values are as follows:
-6 -1 o
t-butyl hydroperoxide Kd = 9.27×10 s at 101 C
-7 -1 o
t-butyl peroxide Kd = 6.82×10 s at 100 C
The ozonized polymers give the following values :
77 -4 o
EVA Kd = 8.87×10 at 98 C
78 -3 o
PVC Kd = 2.14×10 at 90 C
75 -3 o
PE Kd = 10 at 110 C
76 -3 o
PVDF Kd = 8.8×10 at 90 C
This may be explained by the fact that reagents such as DPPH present dur-
ing grafting or titration accelerate the decomposition of the peroxidic functional
groups, thus increasing the dissociation constant.
The activation energies are about three times lower than for classical initi-
ators and the efficiency factor calculated for ozonized EVA remains low, that is
o o
10% at 70 C and 30% at 130 C. Steric and chain configuration hindrances can
account for these low f values.
3.4 THE GRAFTING METHODS
Polymers activated by ozone are directly grafted following activation or after a

lapse of time, considering that they keep their activity long after. In this section,
we describe the main grafting techniques.
3.4.1 Grafting in solution
This grafting technique is performed in solvent which permits to decrease vis-
cosity and favors swelling of the polymer powder, thus enhancing the monomer
penetration into the core. In some cases, grafting is performed in the monomer
alone which then plays a role of solvent. For the most part, reactions are carried
out in a reactor with an efficient mechanical stirring, under a stream of nitrogen
and with a heating jacket.
3.4.2 Grafting in bulk
The bulk grafting technique consists of swelling the activated polymer in the
monomer used for grafting (if monomer is liquid), or mixing polymer with the
monomer powder when the monomer is solid. The mixture is then introduced in
a brabender-type, batch mixer which is thermally regulated and has efficient
mixing characteristics. The reaction is followed by recording torque. The reac-
tion is consider complete when the torque stabilizes. Figure 1 gives examples of
recorded values during styrene grafting reaction onto polyethylene. The influ-
ence of the temperature on the grafting rate can be observed from Figure 1. A
higher temperature leads to a quick stabilization of torque, indicative of a fast
grafting reaction. It is also important to notice that the mixing of ozonized poly-
ethylene alone or non-ozonized polyethylene mixed with styrene does not cause
any rapid increase of torque which suggests that the recorded torque values ex-
79
press the grafting reaction.
3.5 ANALYSIS OF PRODUCTS
The products of the grafting phase contain grafted copolymers mixed with
homopolymer formed by thermal polymerization or resulting from the thermal
decomposition of hydroperoxides. It is important to separate these components
to characterize the products obtained. It is thus necessary to purify and separate
the grafted copolymers.
The purification of grafted copolymers is generally performed by fractional
precipitation of the reaction mixture in a pair of solvent and non-solvent of the
polymers. As an example, we will mention the grafting reaction of PVDF by

methylmethacrylate where a crude reaction product is purified as follows:
The soluble fraction contains homopolymers (whatever their origin: ther-

mal or hydroperoxides) as well as copolymers, highly grafted by PMMA, with a
solubility close to that of pure PMMA, and the insoluble fraction contains the
grafted copolymers. It should be noted that this operation may be repeated if an
extreme purity of the products obtained is required, but single operation leads to
35
sufficiently pure copolymers.
The analysis made by various researchers working on this topic include
several methods. Grafted copolymers are characterized by their molar or mass
grafting rates. The latter is generally obtained from the elemental analysis of
76
the elements present in the copolymer. Serdani et al. determined fluorine and
carbon in PVDF copolymer grafted by PMMA. The general structure of their co-
polymer is as follows:
Figure 1: Styrene grafting reaction followed by measurement of torque in the batch mixer.
They suggest that the molar ratio of the two types of sequences is given by
the following formula involving the determined fluorine amounts:
3800 − 64%F
x/y=
100%F
The same calculation can be done from the results of determination of the
carbon amount. The molar grafting rate is obtained by the following formula, de-
rived from the previous one:
3800 − 64%F
T=
3800 + 36%F
This method has been the most frequently applied to determine the graft-
31,33,67
ing rates, and it is faster and more practical. The results can also be ob-
tained from NMR for polymers soluble in usual solvents.
33
Robin et al. applied gel permeation chromatography, GPC to P(VCgSt) co-
polymers prepared by this method (Figure 2). It was demonstrated that PVC
ozonization leads to a decrease in the molecular weight of polymers and an in-
crease in polydispersity, which is in full agreement with what we have seen in
the first part of this chapter. After grafting, a clear increase of the copolymer mo-
lecular mass is noticed, thus indicating that the grafting has been achieved. The
1
H NMR of the copolymer gives a styrene amount in the range of 13%.
Figure 2. GPC chromatograms of various samples based on PVC
The copolymers structure characterization was improved by various au-

75
thors. Sarraf et al. determined the percentages of peroxides and
o
hydroperoxides obtained after polyethylene ozonization at 60 C by measuring
the molecular masses before and after ozonization and assuming that each per-
oxide leads to a decrease of the molecular mass after decomposition and that hy-
dro-peroxides do not induce any variation of mass. They estimated that their
samples contained 60% peroxides and 40% hydroperoxides. Knowing this ratio,
the molecular mass of the ozonized polymer, and the total number of peroxide
and hydroperoxide sites, they derived a formula which allows one to calculate
the number of grafting sites in the molecule, that is, one or two sites per chain.
When the MMA grafting rate was calculated from the elemental analysis of C
and H, the length of the grafts was estimated to be between 20 and 40 links, de-
pending on the amount of monomer present during grafting. The authors of this
work notice that the number of active sites in the chain (1 or 2) has little influ-
ence on the chain length of the grafter, but the amount of MMA added plays an
important role.
77
With the same purpose, Fargere et al. demonstrated that DPPH titrates
only hydroperoxides without taking into account peroxides which can be esti-
mated from the determination of the cleavage numbers in a chain. These au-
thors established that only peroxides are responsible for the chain cleavages
and thus for the decrease of the molecular mass of the polymer measured by
75
GPC. On the contrary, Sarraf et al. show that ozonized PE contains 4% of per-
oxides and 96% of hydroperoxides with a peroxide number in a chain of about 1
for 10 hydroperoxides. They attempted to measure the kinetic length of the
grafts by assuming that grafts have the same chain length as homopolymers
formed in the medium. This assumption is necessary because the attempts by
66
Brugière, who tried to cleave the ether bonds, linking the grafts to the polymer
backbone, to measure their length, proved to be unsuccessful. Beginning with
this assumption and measuring the molecular weight of homopolymers by GPC,
these authors found the graft lengths between 200 and 600 monomer units (very
large). These high values are not in accordance with the theoretical values cal-
culated from the kinetic laws involving termination reactions by classical re-
combination for vinyl monomers.
The authors deduce that the transfer side-reactions occur during the graft-
ing, thus limiting the grafts lengths.
3.6 EXAMPLES OF APPLICATIONS OF GRAFTED COPOLYMERS
3.6.1 Particular case of plastics recycling

Number of grafted copolymers were prepared by the ozonization and grafting
technique. Polyolefins, PE and PP, were grafted with fluorinated acrylates to
obtain fluorine content between 11% and 25% that highly modified their proper-
31
ties. Likewise, PE and PVDF have been grafted with various acrylic monomers
with the aim to improve their adhesive properties. From these products, it has
been possible to attach composite materials, based on epoxy resins, to PE or
35,80,81
PVDF. The main copolymers prepared according to this technique are
summarized in Table 1.
Such a procedure may also be used to prepare ion exchange membranes.
32
Thus, El Midaoui et al. grafted 2-dimethylaminoethyl methacrylate or acrylic
acid onto polyethylene in order to decrease the electric resistance of this copoly-
mer and to increase the transport of ions.
Table 1: Examples of grafted copolymers prepared by ozonization
Copolymer Mass grafting rate (%) Reference

P(E g 2-hydroxyethyl
10 80
acrylate)
P(E g acrylic acid) 20 80
P(E g glycidyl methacrylate) 10 80
P(E g vinyl acetate) 25 80
P(VDF g St) 17 35
P(VDF g AA) 12 35
P(VDF g MAGLY) 50 35
P(VDF g MMA) 29 35
The ozonization technique has also been used to process plastic wastes
from individual materials or their mixture. Various approaches to the problem
were realized.
4 POLYMER BLENDS AND ALLOYS
Along this study, we have tried to obtain polymer blends without previous sepa-
ration of the mixture constituents. For this purpose, we prepared
compatibilizers, namely grafted copolymers, for which the effect on incompati-
82
ble polymers has been known for a long time. It is well known that either
diblock, grafted or multiblock copolymers, when added in small amounts to in-
compatible mixtures, improve properties such as:
• phase dispersion
• strength at break
• elongation at break
• impact resistance.
But this type of copolymers is often difficult to prepare especially diblock
copolymers, and their use is generally limited to virgin materials. We developed
the technique of preparation of grafted copolymers by ozonization, simple to con-
duct and inexpensive, which can be applied to material recycling.
The technique consists of preparation of copolymer separately and adding
it to the mixture during its processing. In all examples described below, 5% or
10% copolymer was added to compositions mixed under strong shear in a

Brabender-type hot-mixing cell. Then, the mixtures were molded by compres-
sion into sheets from which specimens for mechanical testing were cut out by
stamping. Morphological studies of the phases were done using a scanning elec-
tron microscope for fracture faces of the samples immersed in liquid nitrogen.
4.1 PE/PVC BLENDS83,84
Sarraf et al. studied PE/PVC blends compatibilized by copolymers prepared by

the ozonization and grafting technique. Tensile strength data are available from
this study for 50/50 wt% composition mixtures, containing 5 wt% copolymer (Ta-
ble 2).
Table 2: Tensile strength at break versus amount

of compatibilizer added to PE/PVC blends
7
Copolymer concentration in PE/PVC blend (phr) σR (10 Pa)
0 0.60
2.5 1.07
5 1.15
10 1.17
15 1.16
We notice that the best performance is obtained when 5 wt% copolymer is

added. For lower copolymer concentration, the performance is gradually im-
proved.
Copolymers prepared by grafting of monomers after ozonization always
show some amounts of homopolymers due to the thermal or hydroperoxide initi-
ation. Authors have tested the compatibilizing effect of copolymer, after or be-
fore purification. Table 3 summarizes our results for PE/PVC blends to which 5
wt% of P(EgMMA) copolymer, grafted with 50% PMMA, was added.
The results show that the addition of copolymer considerably improves the
mechanical properties of the mixtures, and that the use of a non-purified copoly-
mer leads to the same performance as purified product. The purification step is
thus unnecessary in this kind of application. It should also be noticed that the
strength at break and the modulus are the two important characteristics. The
elongation at break remains almost constant for this type of copolymer, and it
even decreases sometimes, thus indicating that the mixtures acquire a fragile
behavior (Figures 3 and 4).
Table 3: Mechanical performance of PE/PVC blends versus composition

7 7
Ratio σR (10 Pa) ε (%) E (10 Pa)
PE PVC a b c a b c a b c
100 0 0.80 0.80 0.80 238 238 238 9.7 9.7 9.7
80 20 0.64 0.87 0.88 54.2 57.0 56.0 12.6 12.7 12.8
60 40 0.62 1.07 1.06 28.4 20.3 19.7 13.4 23.5 22.4
50 50 0.60 1.25 1.26 15.5 12.7 13.3 24.3 33.7 33.8
40 60 0.67 1.38 1.35 9.30 7.30 7.50 30.9 45.2 43.7
20 80 1.27 1.88 1.85 3.40 7.60 7.30 50.3 70.2 70.2
0 100 4.80 4.80 4.80 7.60 7.60 7.60 106.3 106.3 106.3
a - without addition of copolymer, b - with 5% purified copolymer, c - with 5% non-purified copolymer
85
The processing was reproduced on a twin-screw extruder by B. Leseur,
who optimized the machine parameters for compatibilized mixtures and con-
firmed the results obtained by Sarraf et al. The water absorption of the prepared
alloys was tested by immersion of the materials in boiling water. Since the mix-
ture interphase was improved by the use of compatibilizer, much less water ab-
sorption was observed for compatibilized mixture (0.12% instead of 0.25%)
which confirms the effect of the copolymer on the interfacial adhesion.
4.2 PE/PS BLENDS
Mixtures of polyethylene and polystyrene were blended using P(EgS) copoly-

mers having mass grafting rate of 31% and an average molecular mass of about
20,000 (an average one graft per chain of a molecular mass of 6,000). For 50/50
86
mixtures, Sarraf et al. have determined the optimum amount of their copoly-
mer required to compatibilize mixture. The results are presented in Table 4.
Table 4: Tensile strength at break of PE/PS

blends versus concentration of compatibilizer
7
Copolymer per 100 g of PE/PVC blend σR (10 Pa)
0 0.79
1.5 1.20
2.5 1.36
5 1.35
10 1.26
Figure 3 : Variation of the PE/PVC mixtures strength at break, sR. - no compatibilizer, l - with 5
wt% of non-purified compatibilizer.
It should be noticed that very low amounts of the copolymer are sufficient
to compatibilize blend, and that above 2.5 wt%, the strength at break of the mix-
ture decreases considerably.
Figure 4 : Variation of the PE/PVC mixtures elongation at break, eR. - no compatibilizer, l - with
5 wt% of non-purified compatibilizer.
Mixtures without compatibilizer exhibit mechanical performances well

within the arithmetic middle terms of the mixture compositions. This difference
is the largest for the 50/50 mixture. When 2.5 wt% copolymer are added to the
mixture, the mechanical performance is considerably increased, even if it is infe-
rior to the arithmetic middle terms, and especially the value of the strength at
break is improved by a factor 2 for the 50/50 composition. Concerning the elon-
gation at break, Figure 5 shows a peculiar phenomenon.
Table 5: Mechanical performance of PE/PS mixtures depending on composition

7 7
Ratio σR (10 Pa) ε (%) E (10 Pa)
PE PS a b a b a b
100 0 0.80 0.80 238 240 9.72 9.7
80 20 0.74 0.89 60 77 15.2 15.9
60 40 0.65 1.05 15 24 22.7 25.5
50 50 0.79 1.30 5.9 15.3 50.1 48.2
40 60 1.07 1.66 6.1 8.6 64.1 80.5
20 80 1.53 2.20 12.3 3.4 126 111
0 100 3.21 3.21 4.37 4.3 128 128
a - without addition of copolymer, b - with addition of copolymer
Table 6: Mechanical performance of compatibilized

and non-compatibilized blends
6
Blend composition σR (10 Pa) ε (%)
Standard PVC 35 82
Standard PS 25.6 2.7
PVC/PS = 50/50 26 2.7
PVC/PS/coplymer = 50/50/10 36 4.1
For mixtures without compatibilizer, containing between 0 and 30 wt% PE,

we notice that ε value reaches a maximum for 20% PE. Addition of copolymer
does not improve elongation which even decreases. Above 30 wt% PE, the in-
verse effect is observed. The curve without copolymer exhibits a minimum
around 50 wt% PE, when the elongation for compatibilized blend increases and
it is always higher than elongation for material which was not compatibilized.
Addition of P(EgS) does not modify elongation at break of materials in propor-
tion to the amount of compatibilizer added, contrary to what is observed for
85
strength at break. Leseur optimized the processing of similar blend on indus-
Figure 5. Variation of the PE/PS mixtures elongation at break. - no compatibilizer, l - with 5

wt% compatibilizer.
trial extruder (speed of rotation, extrusion temperature) and provided evidence

for a good cohesion of phases by calculating the percentage of water recovery for
various mixtures. He shows that compatibilized mixtures absorb twice less wa-
ter than non-compatibilized mixtures and that the phases size is improved.
Using scanning electron microscopy, he also shows a decrease in the nodules
size as well as a better dispersion of these nodules in the continuous phase, simi-
86
lar to indicated by Sarraf et al.
Photographs 1 and 2. Morphology of PVC/PS alloys. Magnifications: 2,000x and 10,000x.
4.3 PS/PVC BLENDS
The use of P(VCgS) copolymers prepared according to the technique herein de-
33
scribed was considered by Robin et al. who tested the 50/50 mixtures contain-
ing 10% grafted copolymer. The results obtained are included in Table 6.
The authors show that the addition of grafted copolymer affects the
strength at break, and, in a lesser proportion, the elongation at break. As previ-
ously mentioned, the copolymer purification after synthesis is irrelevant for the
performance, and, actually, the best performance is reached with non-purified
copolymers. SEM micrographs (Photographs 1,2,3, and 4) show evident im-
provement in the phases dispersion and a decrease of the nodules size by a factor
5 for compatibilized blends.
Photographs 3 and 4. Morphology of PVC/PS/P(CVgS) alloys. Magnifications: 2,000x and 10,000x.
4.4 PS/PVC/PE BLENDS
PE/PVC/PS ternary mixtures containing PEgPS and PEgPMMA copolymers

85
were studied for 1/1/1 mixtures of homopolymers containing 5 wt% of each of
the copolymers to confirm the results presented in the previous studies, namely,
the compatibilizing effect of copolymers prepared by ozonization.
Fractographs obtained for mixtures and water absorption measurements
confirm the improvement of the interphases quality and the decrease in the
phases size.
Table 7: Mechanical properties of PE/PVC/PS ternary mixtures

7
Modulus (10 Pa) Strain (107 Pa)
PVC/PE/PS = 100/100/100 114 1.27
PVC/PE/PS/P(EgS)/P(EgMMA) = 100/100/100/5/5 194 1.88
Figure 6. Changes of strength at break in PVDF/PS blends. - no compatibilizer, l - with

compatibilizer.
4.5 PVDF/PS BLENDS

35
Within the framework of this study, Serdani et al. have tested P(VDFgS) copol-
ymers effect on the performance of the mixtures and they showed that the opti-
mum amount of copolymer is 8 wt%. The change of the mechanical properties as
a function of the mixture composition was studied in detail and is summarized
in Figure 6. We note that the copolymer efficiency is at the maximum for 50/50
composition, offering 150% increase of the strength at break. For compositions
close to 80/20 and 20/80, we observe that the copolymer influence is already ef-
fective.
4.6 EVA/PS BLENDS
77
For the purpose of their studies, to determine f and Kd values, Fargère et al.
prepared P(EVAgS) grafted copolymers that were used as compatibilizers for
EVA/PS blends. The strength at break of 50/50 mixtures of homopolymers in-
creases from 65% to 79% and the elongation at break is increased by a factor 8.
SEM explains this improvement. The basic mixtures without compatibilizer
have the properties of a highly heterophasic system without phases cohesion.
The compatibilized mixtures are more homogeneous without phase discontinu-
ity. This confirms the compatibilizer effect brought by the EVA/PS grafted copol-
ymers.
5 MONOMATERIAL WASTE SOURCES
APPLICATION TO POLYMER-BITUMEN BLENDS
5.1 POLYETHYLENE WASTES
The treated wastes were essentially ground polyethylene films coming from
packing bags, pallet shrinks, greenhouse films, and all other kinds of wastes.
These materials were ozonized in a fluidized bed, according to the technique pre-
viously described and added to bitumen at the rate of 10.5 wt%. The mixing of bi-
tumen and polymer mixture is carried out under strong stirring (10,000 RPM).
o
The studies show that mixing at 180 C for one hour and thirty minutes is neces-
sary to obtain an suitable mixture. The performance of mixtures containing bi-
tumen (180/220)/polyethylene (fluidity indices IF of 4 to 70) are included in
Table 8. It should be noticed that the penetrability is highly influenced by poly-
ethylene incorporation (sometimes even by a factor 2) and that the use of
ozonized polyethylene slightly increases penetrability compared to its
non-ozonized homologous. The ring-and-ball temperatures, representing the
thermal properties of binders rapidly increase when ozonized polyethylene is

added. It should be noticed that ozonized polyethylene contributes to higher in-
crease of ring-and-ball temperature, thus offers improvement of thermal prop-
erties of bitumen. The ozonized polyethylene also improves bitumen fluidity,
87,88
compared to bitumens containing a standard polyethylene.
Table 8: Main physical characteristics of bitumen/polymer blends

Fraas Penetrability Ring-and -ball Viscosity
o o
point at 12.5 C temperature at 160 C
o o -3
( C) (1/10 mm) ( C) (10 Pa.s)
Standard bitumen - 40 38 100
Non-ozonized bitumen + PE (IF=70) - 24 80 850
Ozonized bitumen + PE (IF=70) - 20 90 800
Non-ozonized bitumen + PE (IF=20) -11 35 - 1315
Ozonized bitumen + PE (IF=20) -16 25 - 390
Non-ozonized bitumen + PE (IF=4) -15 17 - 1920
Ozonized bitumen + PE (IF=4) -18 15 - 1300
The use of ozonized polyethylene also improves low temperature proper-

ties, compared to non-ozonized polyethylene, which is expressed by significant
o
differences (3 to 5 C) in the Fraas point.
The X fluorescence investigation of the materials, prepared in this manner,
shows that the dispersion of the polymer inside the bitumen is much more pro-
nounced in the case of activated polymer, regardless of the polymer grain size in-
troduced during the mixture preparation. All these results show the advantages
of the polymer activation. Below, we attempt verification of reasons of this im-
provement.
In order to understand this phenomenon and establish if a chemical reac-
tion occurs between the ozonized PE and the bitumen, PE-bitumen coupling
products were isolated. For this purpose, an ozonized bitumen/PE binder was
purified by carbon tetrachloride washing in order to eliminate the bituminous
fraction. The recovered polymer was crystallized in hot toluene and analyzed.
The product isolated in such a manner exhibited characteristics typical for poly-
ethylene and was studied by GPC chromatography (Table 9).
Table 9: Molecular weight and polydispersity of the ozonized

polyethylene deactivated on bitumen
*Mp *Mn I
Standard polyethylene 123,200 13,700 9
Standard ozonized polyethylene 73,100 6,100 12
Ozonized polyethylene deactivated on bitumen 90,100 15,800 5.7
Standard bitumen 1,390 430 3.3
*according to PS calibration
The results show that ozonization decreases the molecular weight of poly-
ethylene molecules. The molecular mass of the ozonized product is approxi-
mately twice lower than that of initial product. Moreover, the polydispersity
index rapidly increases, which indicates that chains, broken during the
ozonization, have different sizes. When we examine the results obtained with
the ozonized sample deactivated on bitumen, we notice that the molecular mass
increases and is in the range of the standard polyethylene molecular mass. This
increase can only be due to two phenomena:
• either free radicals coupling reactions of the following types occur:
RO • + R′ O • → ROOR′
RO • + H • → ROH
RO • + R′ • → ROR′
Only the two last reactions lead to stable products.

• or a deactivation on the bitumen occurs with grafting of bituminous prod-
uct onto the polymer backbone.
The probability for the second hypothesis is higher because the reaction
medium mainly contains bitumen and the Ro∴φ9 free radicals have far more
chances to deactivate on bitumen. Furthermore, molecules constituting the bi-
tumen have low molecular weight (< 1500), as verified by GPC analysis
(Table 9). GPC data also indicate that the number of grafts, reacted with
ozonized PE, is large because Mn varies from 6100 to 15800.
The analysis was completed by proton NMR measurements in a solid state,
and it was possible to establish that aromatic and heteroaromatic structures
were grafted onto the ozonized polyethylene, thus pointing at deactivation of

polymer by some bituminous fractions. A better compatibility of the polymer
with the bitumen and thus better material performances follow.
Gravel-bituminous mixes were also studied with various types of gravels
and their mechanical performances were determined for standardized speci-
88,89
mens. The simple and “Brasilian” type compression resistance were mainly
o o
studied at 18 C and 50 C.
Table 10: Simple compression at 18 and 50oC and “Brasilian” measurements for
various gravel-bituminous mixes
o o
18 C 50 C Brasilian
5 5 5 Ratio*
(10 Pa) (10 Pa) (10 Pa)
Standard gravel-bituminous mixes without PE
83 17 20 4.8
(5.5% bitumen)
PE-based gravel-bituminous mixes (IF=70; Ø<800 µm) 117 36 26 3.25
Ozonized PE-based gravel-bituminous mixes (IF=70;
129 37 26 3.48
Ø<800 µm)
Finely ground recovered PE-based gravel-bituminous
136 46 - 3
mixes (Ø<2.8 mm)
Finely ground and ozonized recovered PE-based
150 50 34 3
gravel-bituminous mixes (Ø<2.8 mm)
Coarsely ground recovered PE-based gravel-bituminous
125 51 27 2.4
mixes (Ø<6 mm)
*ratio of compression at 18 and 50oC, Rc
In simple compression testing, the specimen is impacted along the height
and the force measured is related to the specimen cross-section. In “Brasilian
test”, the specimen is impacted at the edge and the force related to the height
and the diameter of the specimen. The results are reported in Table 10. From
these results, we can conclude that the addition of PE in gravel-bituminous
mixes improves their mechanical properties, compared to the standard
gravel-bituminous mixes without polymer. If we compare gravel-bituminous
mixes to PE and their homologoes to ozonized PE, we can see that the perfor-
o
mance of the latter is better, especially at 18 C. We also notice that the perfor-
mance of a finely ground PE is superior to samples obtained with a coarsely
o
ground polymer. This difference is considerable at 18 C.
o o
Resistance in simple compression, at 18 C and 50 C, have been compared.
The Rc 18/50 ratio is representative for the thermal susceptibility of the
gravel-bituminous mixes. If the gravel-bituminous mix is thermally suscepti-
ble, the Rc value is higher due to more rapid decrease in compression resistance
o
at 50 C. In this study, the Rc 18/50 ratio values are close in the range of 3.25 to
2.4. For all gravel-bituminous mixes containing polyethylene, we obtain a value
lower than that obtained with gravel-bituminous mixes made out of bitumen.
This proves that the thermal susceptibility of gravel-bituminous mixes with
polyethylene is lower and that the thermal properties of these compositions
were improved.
Concerning the “Brasilian” type tests, we find that addition of polymer
highly improves the compression resistance, compared to traditional gravel-bi-
tuminous mixes.
The flexural modulus of gravel-bituminous mixes was measured at various
temperatures and the results are given in Table 11. The flexural modulus of
gravel-bituminous mixes, based on ozonized PE, is superior to their
non-ozonized precursors.
Table 11: Flexural modulus of gravel bituminous mixes based on bitumen and
ozonized polyethylene
o o o
E at 10 C E at 15 C E at 25 C
(GPa) (GPa) (GPa)
Reference gravel-bitumen mix 0.088 0.055 0.016
Gravel-bituminous mix based on ozonized PE (IF=70) 0.407 0.235 0.182
Gravel-bituminous mix based on non-ozonized PE (IF=70) 0.191 0.051 0.043
Gravel-bituminous mix based on recovered ozonized PE 0.462 0.253 0.130
Gravel-bituminous mix based on recovered non-ozonized PE 0.390 0.113 0.100
The binder adhesivity to gravel was also measured, particularly for quartz-
ite, (the mixing with bituminous binders is difficult). Gravel-bituminous mixes
o
with various binders were maintained at temperature 60 C in a thermostated
bath and the percentage of the minerals which were not coated by bitumen
quantified. The results in Table 12 show that binders containing ozonized poly-
ethylene adhere much better than binders based on non-activated polyethylene
or bitumen alone.
Table 12: Adhesivity of binders to gravel

Residual adhesion after wa-
ter
o
immersion (60 C, 24 h) (%)
Quarzite + bitumen 180/200 20
Quarzite + bitumen 180/200 + PE (IF=70) 20
Quarzite + bitumen 180/200 + ozonized PE (IF=70) 100
Quarzite + bitumen 180/200 + recovered PE 20
Quarzite + bitumen 180/200 + recovered ozonized PE 100
5.2 CROSSLINKED ETHYLENE-VINYL ACETATE, EVA, COPOLYMER WASTES
This work was conducted with the firm MALET from Toulouse. We have used
the ozonization technique to decrosslink an ethylene-vinyl acetate copolymer,
EVA, which is an uncommon operation. Wastes coming from various industries
were ozonized in carbon tetrachloride. EVA ozonization is difficult in the solid
state because the waste is damaged by carbonization. In a solvent such as CCl4,
we first observe a dissolution of O3 then a reaction of the gas with the crosslinked
polymer. The authors optimized this reaction according to experimental design,
in which numerous parameters were varied. From structural analysis of model
polymers, it can be anticipated that vinyl sequences are only little affected by
the action of ozone. Essentially, the ethylenic segments are damaged leading to
the decrosslinking of the polymer.
The decrosslinked polymers were incorporated into bitumen in order to im-
prove its mechanical and physical performances. EVA-bitumen mixtures have
already been perfected by road constructors. They include virgin EVA, not
crosslinked, but are seldom used because of a high cost of EVA. Using this type of
waste opens new prospects in this field. The following observations were made
from results obtained:
• The best properties of the bitumen-decrosslinked polymer binders are ob-
tained for wastes ozonized for 90 minutes at 47oC. In other conditions, ei-
ther insufficient dissolution of the polymer in the bitumen, or too complete
dissolution of the polymer, accompanied by decrease in binder perfor-
mance, occurs.
• The polymer should be added to the bitumen as a powder as finely ground

as possible. On the contrary, during the peroxide and hydroperoxide de-
composition, a crosslinking of the polymer occurs, rather than its deactiva-
tion by bitumen (grafting).
• The binders prepared according to this technique are stable and do not de-
cant as time passes, which proves that the polymer is completely miscible
in the medium.
• The binders obtained according to this technique have properties very close
to those obtained for mixtures containing non-crosslinked pure EVA.
6 CONCLUSIONS
Ozonization constitutes an interesting way to deactivate polymers. Easy to pro-
cess, it allows one to obtain reactive functionality which can be used to carry out
grafting reactions leading to copolymers exhibiting specific properties (adhe-
sion, ion exchange, compatibilizing effect, etc.). Thus, numbers of polymers can
be modified, either by reactions carried out in solution, or by reactions with pow-
dered polymers. Among the possible applications of this procedure, plastics re-
cycling has been most widely studied. Effectively, by ozonization, it is possible
to:
• prepare grafted copolymers which can be used as compatibilizers for im-
miscible materials. Within this framework, many mixtures have been
studied and in all cases, the compatibilizers prepared have shown a clear
improvement of the mechanical performance of mixtures studied
• decrosslinked polymers. This type of reaction is not usual and the proce-
dure opens the way to processing of crosslinked polymer wastes.
• activate polymers such as polyethylene in order to improve their compati-
bility with other macromolecular materials such as bitumen.
• the installations are very simple (ozone generators are simple in concept
and use) to design, it is undoubtedly an interesting way of reclaiming plas-
tic wastes.
7 REFERENCES
1. S. Zhiqiang and W. Baker, J. Appl. Polym. Sci., 44, 2167 (1992).
2. A. Simmons and W. Baker, Polym. Commun., 31, 20 (1990).
3. M. Lambla, J.Druz, and N. Satyanarayana, Makromol. Chem., 189, 2717 (1988).
4. M. Glotin, R. Parsy, and P. Abadie, European Patent 0342066 (1989).
5. J. Rösh and R. Mülhaupf, Makromol. Chem. Rapid. Com., 14, 503 (1993).
6. D. Angier and W. Watson, I.R.I. Trans., 33, 22 (1957).

7. G. Smets and M. Claesen, J. Polym. Sci., 7(3), 289 (1951).
8. P. Allen, J. Downer, G. Hasting, and H. Melville, Nature, 177, 910 (1956).
9. G. Mino and S. Kaiserman, J. Polym. Sci., 31, 242 (1958).
10. K. P. Adwait, C. P. Nrusinga, and L. Padma, J. Appl. Polym. Sci., 27, 1873 (1982).
11. D. Hill, A. McMillan, J. H. O’Donnel, and P. Pomery, J. Makromol. Chem. Makromol.
Symp., 33, 201 (1990).
12. W. Cooper, Chem. Ind. (London)., 407 (1953).
13. L. Valentine and C. B. Chapman, Ric. Sci., 25, 278 (1955).
14. W. Hahn and A. Fisher, Makromol. Chem., 21, 77 (1956).
15. J. F. Rabek in Mechanism of Photophysical Processes and Photochemical
Reactions in Polymer, Wiley, New York (1987).
16. D. R. Randell in Radiation Curing of Polymers, Royal Society of Chemistry,
Cambridge (1991).
17. F. Zhengo and B. Ranby, Angew. Makromol. Chem., 195, 17 (1992).
18. R. B. Mesorbian, Proc. U. N. Inter. Conf. Peaceful Uses of Energy, 2nd, Geneva, 29,
196 (1958).
19. H. Hani and T. Uchino, Japan Patent 17637 (1962).
20. American Machine and Foundry, Brit. Patent 872217 (1962).
21. L. L. Kocherginskaya and N. D. Rozenblum, Chem. Abst., 64, 9893 (1966).
22. T. Cohen, French Patent 2 641 539 (1990).
23. J. Blaise, French Patent 2 476 657 (1980).
24. G. V. Shultz, G. Henrici, and S. Olive, J. Polym. Sci., 17, 45 (1955).
25. U. S. Patent 2 947 719 (1954).
26. A. Masakazu, Japan Patent Kokai 7750 390 (1977).
27. R. J.Ceresa, U. S. Patent 3 110 695 (1963).
28. R. B. Mesrobian and H. Mark, Textile J., 23, 294 (1953).
29. G. Natta, E. Beati, and F. Severini, J. Polym. Sci., 34, 685 (1959).
30. A. Holmstorm and E. M.Sorvik, J. Polym. Sci., Polym. Chem. Ed., 16, 2555 (1978).
31. B. Boutevin, J. Mouanda, Y. Pietrasanta, and M.Taha, Eur. Polym. J., 21(2), 181
(1985).
32. A. El Midaoui, T. Sarraf, C. Gavach, and B. Boutevin, J. Appl. Polym. Sci.,42(9), 2551
(1991) and J. Polym. Sci., B29,705.
33. B. Boutevin, Y. Pietrasanta, J. J. Robin, and T. Pollet, Eur. Polym. J., 24 (10), 953
(1988).
34. B. Boutevin, Y. Pietrasanta, and J. J. R. Robin, Eur. Polym. J., 27(8), 815 (1991).
35. B. Boutevin, J. J. R. Robin, and A. Serdani, Eur. Polym. J., 28(12), 1507 (1992).
36. A. Chapiro, J. Polym. Sci., 48, 109 (1960).
37. A. Chapiro, M. Magat, and J. Sebban, French Patent 1 125 537 (1956).
38. A. Chapiro and M. Foex, Eur. Polym. J., 8, 1301 (1972).
39. K. Ito, N. Usami, and Y. Yamashita, Macromolecules, 13, 216 (1980).
40. Y. Yamashita, Y. Tsukahara, and K. Ito, Polym. Bull., 5, 335 (1981).
41. Y. Chujo, J. Shishino, Y. Tsukahora, and Y. Yamashita, Polym. J., 17, 133 (1985).
42. M. Niwa, T. Hayashi, and T. Matsumoto, J. Macromol. Sci. Chem., A23(4), 433 (1986).
43. Y. Yamashita, K. Ito, H. Mizuno, and K. Okada, Polym. J., 14, 255, (1982).
44. M. Akashi, I. Yamashita, and N. Miyauchi, Angew. Makromol. Chem., 132, 81 (1985).
45. Y. Chujo, K. Murai, Y. Yamashita, and Y. Okomura, Makromol. Chem., 186,1203
(1985).
46. B. Boutevin, Y. Pietrasanta, M. Taha, and T. E. Sarraf, Makromol. Chem., 184, 2401
(1983).
47. C. Bonardi, B. Boutevin, Y. Pietrasanta, and M.Taha, Makromol. Chem., 186, 261
(1985).
48. Y. Nitadori and T. Tsuruta, Makromol. Chem., 180, 1877 (1979).
49. T. Nishimura, M. Maeda, Y. Nitadori, and T. Tsuruta, Makromol. Chem., 183, 29
(1981).
50. C. C. Schubert and R. N. Pease, J. Amer. Chem. Soc., 78, 2044 and 5553 (1956).
51. V. J. Darvin, L. M. Albright, and H. Wien, Chem. Ber., 92, 958 (1959).
52. M. S. Whiting, A. J. N. Bolt, and J. H. Parish, Adv. Chem. Ser., 77, 4 (1969).
53. N. A. Kleimenov, A. B. Nal’Bandyan, Dokl. Akad. Nauk SSSR., 122, 103 (1958).
54. S. D. Kefeli, S. D. Razumovskii, and G. E. Zaikov; Vysokomol. Soed., 13A 803 (1971).
55. W. Pritzkow and G. Schöppe, J. Prak. Chem., 311, 689, (1969).
56. S. D. Razumovskii, Zh. Organ. Khimii., 3, 789 (1967).
57. S. D. Razumovskii, Zh. Prikl. Khimii., 41, 863 (1968).
58. I. V. Berezin, E. T. Denisov, and N. M. Emmanuel in Okislenie Tsiklogeksana,
Izdatel’stvo GU, Moscow, 1962.
59. R. T. Elworthy, Trans. Roy. Soc. Canada, 16(111), 93 (1922).
60. E. Briner and J. Carseller, Helv. Chim. Acta., 18, 973 (1935).
61. R. Crigee, Chimia, 22, 392 (1968).
62. R. Crigee and P.Günter, Chem. Ber., 96, 1564 (1963).
63. G. Schroder, Chem. Ber., 96, 1564 (1963).
64. P. S. Bailey, G. E. Batterbee, and A. G. Lange, J. Amer. Chem. Soc., 90, 1027 (1969).
65. P. S. Bailey, J. Org. Chem., 22, 98 (1957).
66. S. Bruguière, Ph. D., Synthèse d’une résine échangeuse d’ions à cinétique
rapide sur support polypropylène, University of Lyon (1989).
67. B. Boutevin, Y. Piétrasanta, M. Taha, and T. Sarraf, Eur. Polym. J., 20(12), 1131
(1984).
68. Y. Landler and P. Lebel, J. Polym. Sci., 48, 477 (1960).
69. G. Zeppenfeld, Makromol. Chem., 90, 447 (1996).
70. A. Michel and C. Monnet, Eur. Polym. J., 17, 1145 (1981).
71. J. C. Chien, J. Polym. Sci., Polym. Lett. Ed., 10, 915 (1972).
72. D. L. Allara, C. W. White, and R. L. Meek, J. Polym. Sci., Polym. Chem. Ed., 14, 93
(1976).
73. T. Fargère, M. Abdenadher, M. Delmas, and B. Boutevin, J. Polym. Sci., (submitted
for publication in 1993).
74. A. V. Tobolsky, J. Am. Chem. Soc., 80, 5927 (1958).
75. T. Sarraf, Ph. D., University of Montpellier (1987).
76. A. Serdani, Ph. D., University of Montpellier (1990).
77. T. Fargère, Ph. D., University of Toulouse (1993).

78. A. Michel, E. Castaneda, and A. Guyot, J. Makromol. Sci. Chem., A12, 227 (1978).
79. J. J. Robin, Ph. D., University of Montpellier (1989).
80. B. Boutevin and J. J. Robin, Eur. Polym. J., 26(5), 559, (1990).
81. French Patent 84 06 447 (1984).
82. X. Xanthos, Polym. Eng. Sci., 28, 1392 (1988).
83. B. Boutevin, Y. Pietrasanta, M. Taha, and T. Sarraf, Polym. Bull., 14, 25 (1985).
84. B. Boutevin, Y. Pietrasanta, and T. Sarraf, Angew. Makromol. Chem., 162, 175 (1988).
85. B. Leseur, Ph. D., University of Montpellier (1992).
86. B. Boutevin, Y. Pietrasanta, and T. Sarraf, Angew. Makromol. Chem., 148, 195 (1987).
87. B. Boutevin, Y. Pietrasanta, and J. J. Robin, Eur. Polym. J. 23(6), 415 (1987).
88. French Patent 25 69 416 (1984).
89. B. Boutevin, Y. Pietrasanta, and J. J. Robin, Eur. Polym. J., 23(7), 525 (1987).
90. A. Fahimi, Ph. D., University of Toulouse (1992).
J. Scheirs and G. Camino 167
Effect of Contamination
on the Recycling of Polymers
J. Scheirs
ExcelPlas Australia, P.O. Box 102, Moorabbin 3189, Victoria, Australia
G. Camino
Dipartimento di Chimica Inorganica, Chimica Fisica e Chimica dei Materiali

Università di Torino, Via P. Giuria 7, 10125 Torino, Italy
INTRODUCTION
Contamination in recycling is an extremely broad topic encompassing a di-
versity of forms such as dirt, partially oxidized polymers, printing inks, paper,
pesticides, metals, foil, additives (e.g., fire retardants and antioxidants) and
their transformation products. Contamination in polymers for recycling is ubiq-
uitous; be it soil in LDPE mulch film or metal fragments from aluminum caps in
PET. Furthermore, plastic products are increasingly multicomponent systems
comprising pigments, fillers, labels, glues, laminates and coatings. The poten-
tial for undesirable/antagonistic interactions and the probability of undetected
contamination being present, leading to a reduction of the recycle quality, is
high. Even in curbside collection programs solely for PET and HDPE bottles, col-
lected in plastic boxes, it has been found there can be 15 to 20% contamination
1
by foreign bodies. Table 1 is a comprehensive list of common contaminants
found in recycled polymers.
168 Effect of contamination on the recycling of polymers
Table 1: Common contaminants in recycled polymers
Polymer Recycle source Contamination

PET beverage bottles PVC, green PET, Al, water, glue, oligomers
HDPE milk/water bottles PP, milk residue, pigments, paper, EPS, cork
LDPE greenhouse films insecticides, soil, Ni, oxidation products
LDPE shopping bags paper receipts, printing ink, food scraps
PP battery cases Pb, Cu, acid, grease, dirt
HDPE detergent bottles paper, glue, surfactant, bleach, white spirit
PET photographic film silver halides, gelatin, caustic residues
phenolic circuit boards Cu, tetrabromobisphenol A
LDPE multilayer film ethylene vinyl alcohol, polyamide, ionomer
PVC beverage bottles PET, PE, paper, Al foil, PP
ABS appliance housings polybrominated flame retardants
SBR automobile tires steel wire, fiber, oil extender
LDPE mulch film soil (up to 30%), iron (up to 3% in soil)
SBR - styrene-butadiene rubber; EPS - expanded polystyrene; ABS - acrylonitrile-butadiene-styrene terpolymer
Recycling of polymers is generally considered to be the most environmen-

tally friendly alternative for dealing with the waste disposal problem. Even if
consumers can be persuaded/educated to segregate polymer products prior to
disposal in the municipal waste stream, there will always be the problem of sep-
arating inadvertent mixtures, multicomponent products and adventitious con-
tamination. One process, which is economically attractive and widely used to
separate impurities from a stream which is predominately one polymer, is by us-
ing differences in density. The separation of the HDPE base cup from the PET
beverage bottles is one such example. There is, of course, the alternative not to
separate the plastic waste and to use it in its commingled form. A greater degree
of separation and purification is possible if the polymers are solvated.
Multicomponent plastic products can be separated using mixtures of solvents
2
that allow for selective dissolution.
Finally given that today’s polymer recycle is comprised of materials made

perhaps 20 years ago, this raises the question over the types of additives that are
present and their conversion products after the recycling process. Fire retar-
dants, in particular, are a class of additives that are used in relatively high con-
centrations (e.g. 10-15%) in many polymeric applications from building
products, electrical and electronic systems, transportation vehicles to furniture
and appliance housings. Certain halogenated fire retardants, which have been
banned due to potential toxic hazards, are nonetheless still present in many of
the polymers in use today and may inevitably enter the recycling stream.
CONTAMINATION BY OTHER POLYMERS
PVC in PET
In Europe, the majority of bottles for carbonated beverages and spring wa-
ter are manufactured from either polyethylene terephthalate, PET, or polyvinyl
chloride, PVC. As such, these polymers are the most commonly encountered re-
cycled plastics and provide the main feedstock for the plastics recycling indus-
try. PET and PVC, however, are often found to contaminate each other mutually
by virtue of their very similar specific weights. As a consequence, they cannot be
separated through ordinary physical processes such as water flotation or air
1
elutriation and, therefore must be separated by more sophisticated means.
PVC is not stable at the processing temperatures of PET (i.e., approximately
o
280 C). At these temperatures PVC readily undergoes dehydrochlorination
leading to copious evolution of hydrogen chloride. As a result, many purchasers
of recycled PET specify that PVC be “undetected”. The presence of as little as 100
ppm of PVC as flake in recovered PET can cause serious polymer discoloration of
PET during its drying phase and lead to contamination of the polymer with
3
black specks during extrusion. Yet there are applications where up to several
hundred ppm of PVC is tolerable.
HDPE in PET
Another common polymer used in consumer bottles, especially milk bottles

in the US, is high-density polyethylene, HDPE. Since the consumer does not al-
ways segregate HDPE and PET bottles and because of the high-volume usage
there is a high probability that these two polymers are commingled when arriv-
ing at the recycling plant. These two polymers are incompatible in the melt, and
upon solidification remain as separate phases. The HDPE domains are visually
objectionable, and can cause mechanically weak zones leading to failure in the
4
PET matrix.
PP in HDPE
Polypropylene, PP, is often used to manufacture closures and caps for

HDPE bottles (particularly shampoo, detergent and bleach bottles). Since
HDPE and PP have similar specific gravities they are virtually inseparable by
common physical separation techniques. Sophisticated IR spectrophotometric
separation methods are, however, capable of discriminating between these two
polyolefins. A problem arises during melt processing of HDPE and PP due to the
incompatibility of these polymers. Typical defects that arise due to low level PP
contamination are unfused ‘lumps’ of PP in profiles, poorly-formed parisons in
blow-molding and melt-fracture in extruded films. Furthermore, the addition of
a few percent of PP to HDPE can lower its low temperature impact strength.
This contamination has been thoroughly investigated by FTIR microscope tech-
niques. In contrast, to these findings, work by engineers at Quantum Chemical
Co. (US) claims that up to 10% PP contamination can be tolerated in recycled
5
HDPE.
PET in HDPE
Since PET is a solid at the normal processing temperatures of HDPE,

unmelted PET can quickly plug melt delivery channels and injection nozzles in
4
melt processing machinery.
CELLULOSE in HDPE
Many HDPE and PET bottles (such as milk and beverage containers) have
paper labels which are usually removed during the washing stage in the recy-
cling process. However, residual paper in the form of cellulose fibers have been
found to cause major problems (creation of holes and surface defects) during
blow molding of the reprocessed resin.
MULTILAYER POLYMERS
The advent of coextruded and multilayer films for barrier applications has
complicated the recycling of polymer films due to the presence of polyamides and
EVOH which lack compatibility with the PE or PP recycle.
Table 2: General effect of non-polymeric contamination on polymer recycling
Contaminant Effect on recycling

metals plugs injection nozzles; catalyzes polymer oxidation
paper, fibers ‘blow-outs’ in molded bottles; ‘bubble’ collapse in film extrusion
soil, dirt lowers aesthetic quality of recycle; causes gels and stress concentrations
pigments, dyes unwanted color variation; catalysis of polymer oxidation
water hydrolytic degradation in PET; surface defects in PE
lubricating oil undesirable odors; processing fumes
milk recycle plasticization, lowering impact strength; rancid odor of butyric acid
terephthalic es-
discoloration of PET
ters
hydroperoxides initiates thermal and photooxidative reactions
herbicides toxic fumes present hazard to operators
flame retardants may produce supertoxic compounds
caustic residues fogging of photographic film emulsion on PET
CONTAMINATION BY PACKAGING CONTENTS
Plastic packaging, such as HDPE bottles, are often contaminated by their

contents that have migrated into the polymer or by residues which are difficult
to remove by standard cleaning methods. Often such contamination of the poly-
mer can also arise from non-intended secondary use of plastic containers by the
public for the storage of chemicals. Table 2 summarizes the effects of common
non-polymeric contamination on the recycling of polymers.
6
Hope et al. studied the effect of various contaminants (types of household
industrial chemicals) on the recycling of HDPE bottles. Detergent, bleach, lubri-
cating oil and white spirit bottles were recycled. They found that the contami-
nants generally were absorbed into the walls of the container and were not
removed by thorough washing. The main contaminants in the recycled pellets
were 1-2 wt% of white spirit, 0.7 wt% of hydrocarbon oil, and between 100-500
6
ppm chlorine from the bleach. Another common contaminant in post-consumer
recycle, PCR, HDPE bottles is by residues of motor oil. As a consequence of re-

processing, the oil can impart objectionable odors to the recycled product. Such
oil contamination can be a problem even at levels below the threshold of detec-
tion of sensitive analytical instruments such as gas chromatography-mass spec-
troscopy, GC-MS.
The effect of milk residues on the quality of recycled HDPE from used milk
7
bottles has been studied. It was found that the major component of the rancid
milk was butyric acid which can diffuse into the walls of HDPE bottles. Other
studies have shown that butyric acid contamination in recycled HDPE can lower
its tensile strength and mechanical properties due to an internal
8
plasticization/lubrication effect.
Some organochlorine compounds, especially chlorophenols (e.g.
dichlorophenol, DCP, and trichlorophenol, TCP) have the ability to taint HDPE
at very low levels. In fact, dichlorophenol can produce objectionable odors in
HDPE at levels in the parts per billion range. DCP and TCP are notorious in the
dairy industry for tainting milk bottles where levels as low as 5 ppb can be
readily detected in milk due to their low threshold of taste detection. The typical
scenario by which DCP enters the recycle stream is where post consumer bever-
age containers are used to store herbicides. The residue in one such container
which enters the reuse stream has the ability to taint thousands of unaffected
containers.
Recently there has been effort focused on the recycling of HDPE containers
9,10
which contain pesticide residues. This area poses special problems which are
difficult to overcome since the products formed by pyrolysis or incineration recy-
cling of these containers have not been adequately characterized to date and
may involve the production of highly toxic species.
11
Contamination of PET by foreign substances has been investigated using
health-endangering chemicals to evaluate leaching and migration effects of
these substances in PET. It was concluded that, even with highly toxic sub-
stances such as organochlorine pesticides, the migration into PET is extremely
low and moreover, the leaching of these compounds from the PET is so low (only
a fraction of the migration rate) that the final values are well below the average
11
daily intake specified by the FDA. The incidental contamination of PET has
also been studied using model compounds and diffusion modelling which deter-
mined that the cleaning process associated with conventional PET reprocessing
12
is sufficient to eliminate harmful migratory contaminants.
The studies to date thus show that while HDPE is easily contaminated by
liquid contents, PET is less affected. This can be attributed, in part, to the
oleophilic nature of HDPE and its susceptibility to environmental stress crack-
ing. While PET, when biaxially oriented (as is the case in the stretch-blown bev-
erage bottles), is highly crystalline and generally impervious to the ingress of
migratory compounds. In addition, the relatively high processing temperature
of PET means that any sorbed contamination is likely to break down during re-
processing. HDPE only otherhand, is reprocessed at considerably lower temper-
atures and many toxic compounds could be expected to survive this heat
treatment.
CONTAMINATION BY THE ENVIRONMENT
SOIL IN MULCH FILM
The recycling of mulch film has been actively investigated by Dow Chemi-
cals because the normal practice of disposal by burning it on the fields is envi-
ronmental undesirable. The contamination levels in mulch film make its
recycling particularly challenging. For instance, soil contamination can be as
high as 30-40%. Furthermore, the soil can contain up to 3% of iron which is a
13
polyethylene prodegradant. In addition, it was found that vegetable matter
coming from harvested plants could not be removed during the washing opera-
13
tions. Other contaminants are fumigants (e.g. methyl bromide) and the oxi-
dized fractions of LDPE due to photodegradation of the mulch film.
MOISTURE CONTAMINATION IN PET
Water is a contaminant which induces hydrolytic chain cleavage of PET

14
and as such the polymer should be rigorously dried before melt reprocessing.
o
The drying of recovered PET flakes is performed at temperatures of 160-180 C
in order to lower the moisture content to below 50 ppm which is essential for
11
proper blow-molding of PET in order to avoid MW reduction. Any moisture not
devolatized before the PET becomes molten, rapidly reacts and a surprisingly
small amount of moisture can reduce the viscosity of the polymer to such a level
4
that acceptable bottles cannot be be blown.
CONTAMINATION BY ADDITIVES AND THEIR CONVERSION PRODUCTS
T he polybr ominate d di ph e n y l e t h e rs (de ca -, o ct a -, a n d
penta-bromo-diphenylether, PBDE) are excellent flame retardants which act in
the gas-phase by “flame poisoning”. However, there is concern that these mate-
rials form supertoxic compounds such as polybrominated dibenzofurans, PBDF,
and polybrominated dibenzodioxins, PBDD, during high temperature decompo-
sition as shown in Figure 1. Nevertheless, because of the outstanding efficiency
of PBDE’s they are universally used as flame retardants in most plastics. Such
is their effectiveness that there is currently no equivalent substitute for impart-
15
ing flame retardance to upholstered furniture. Moreover, though the use of
these systems may be curtailed in the near future, the immediate predicament
is that PBDE-containing polymers which were fabricated 10-20 years ago are
now reaching the end of their useful life and entering the recycle stream.
Figure 1. Schematic of reaction pathways to the production of polybrominated dibenzofurans and

polybrominated dibenzodioxins from polybrominated diphenylethers (a common family of flame
retardants for polymers). Note: x+y can equal 5, 8, or 10 implying penta-, octa- or deca-substitu-
tion.
16
A recent paper by Meyer et al. indicates that it is the flame retardants
that determine the recyclability of a polymer. Engineering plastics such as
acrylonitrile-butadiene-styrene which contain PBDE are unsuitable for recy-
cling because of the generation of PBDF and PBDD’s. Moldings containing
16
PBDE must therefore be separated before recycling. The recycling of printed
17
circuit boards has been investigated by Lorenz et al. and it was found that the
combination of the presence of flame retardants (such as, tetrabromobisphenol
A), elevated temperatures and the catalytic effects of copper represent typical
formation conditions for PBDD. Levels of PBDD up to 4.5 ng/g were detected af-
o 17
ter heating the printed circuit boards in an oven at 300 C.
Another important fire retardant is polybrominated biphenyl, PBB, which

is both intrinsically toxic and can form PBDF and/or PBDD in pyrolysis or com-
18
bustion. Today, however, its use is highly limited, but it has been widely used
in the past.
Flame retardant additives can also modify the types of products that are
released during incineration of waste polymers and in this way the flame retar-
dants are contaminants which also affect the quaternary recycling of polymers
(that is, the incineration of polymers to recover heat). For example, the thermal
degradation of polyurethane, PU, can produce a complex mixture of products.
However, in the presence of a common phosphorus-based flame retardant (such
as ammonium polyphosphate), aniline (which is quite toxic) becomes a relevant
19
volatile combustion product.
Another example of additive modification leading to contamination of the
recycle stream involves UV stabilizer based on metal complexes. Low-density
PE greenhouse films often contain UV stabilizers based on nickel complexes
which quench excited states (the so-called nickel quenchers). Whilst, nickel
o
complexes are stable at polyolefin processing temperatures (i.e., 200-250 C),
they degrade at higher temperatures such as those encountered in the process-
o
ing of PET (i.e., >280 C). Therefore, during reprocessing of mixed plastic scrap,
nickel ions can be produced and these, by virtue of their polyvalent nature, can
act as prodegradants and catalyze oxidative decomposition of the polymer.
The pigments used in PE moldings and bags are often based on inexpensive
metal oxides. For instance, the common brown, gray and orange pigments are
based on va rious iron o x i de s a n d h y dra t e s w h i ch a ct a s
o
prooxidants/prodegradants at the high temperatures (i.e., 220 C) encountered
during reprocessing of HDPE. Moreover, green pigments are usually based on
chromium(III) oxide which can readily catalyze the thermooxidative degrada-
tion of HDPE even in trace quantities. Virgin PE is usually adequately stabi-
lized so that these catalytic compounds do not cause in-service degradation of
the polymer, however, with reprocessing the antioxidants are usually consumed
and these pigments may then be able to exert their prodegradant effects.
The recycled polymer can also be contaminated by pigmented constituents
in the feedstock. In the recycling of RDPE bottles by melt reprocessing, a major
20
effort has been directed towards producing a naturally-colored recycle stream.
The major barrier to overcome to reach this end is the removal of the colored bot-
tle caps. Typically blue and red caps are used to differentiate the fat contents of
the milk. Manual removal is too labor intensive and is thus not economical.
Since the weight of the cap is typically 10% of the weight of the bottle, some
recyclers have opted to include the caps into the recycle stream. The effect of this
is that the recycled product has an olive or “dirty” green color due to the presence
21
of red and blue caps. As the caps are generally also PE they cannot be removed
by the float/sink process. Some companies have had limited success removing
the caps by air classification and elutriation techniques, whereby the caps which
are more thicker-walled than the bottles are separated out in a column of air.
There is now a move to produce the caps from PP but this causes other problems
due to its higher melting point. A number of manufacturers have installed color
sorters which rely on electronic optical scanners in combination with high-pres-
sure air jets to selectively remove colored fragments from the recycled material.
Companies such as Simco, Pulsar, ESM, Satake, and Sortex have commercially
available models of such color sorters that are capable of sorting up to 6 tones of
material per hour. Although colored recycle is acceptable for most applications it
is of concern for blow molding of lightly colored bottles.
Color contamination can also occur in PET. Clear PET bottle flake can
sometimes contain some green chips. Perfect manual separation of green from
clear bottles is possible but expensive. Although melt extrusion homogenizes
the polymers and the colors, it takes only 1000 ppm of green PET chips to create
an observable color shift in clear PET.
LDPE film is extensively used for packaging and for the production of shop-
ping bags. These films often contain a fatty-acid amide lubricant (usually
cis-docosenamide) which can be oxidized during thermal reprocessing of LDPE
film. The lubricant readily undergoes cleavage at the unsaturated site to give a
homologous series of aldehydes which have very low odor thresholds. Such con-
tamination imparts to the recycle, a rancid odor which may restrict its applica-
tion potential.
Recycled PET from used x-ray film must be carefully analyzed for contami-
nation if it is to be used for the production of new photographic base film. This is
because caustic residues from the silver recovery process can cause fogging of
22
the photographic emulsion.
CHEMICAL MODIFICATION OF THE POLYMER

Environmental aging and oxidation of polymers can induce chemical modi-
fication of the polymer structure and the production of oxygenated species.
These oxygenated species are in fact “impurities” which can cause sensitization
of the polymer towards further thermal and photooxidation. The most destruc-
tive self generated impurity in many polymers is the hydroperoxide group. Since
hydroperoxides are thermally unstable and photolabile they can readily break
down to initiate a chain of radical reactions. Other impurities are carbonyl
groups which are products of oxidation and can act as chromophores in
photooxidation. For example, greenhouse films, usually based on LDPE, are fre-
quently recycled. These films are usually quite degraded by UV light and are
used to manufacture films for low-grade application. Such films usually have
23
low mechanical strength and low intrinsic stability.
A number of recent papers in the literature have discussed the recycling of
24
fluoropolymers. Although, the reuse of fluoropolymers is relatively scarce com-
pared with commodity thermoplastics, their high cost provides incentive for re-
covery and recycling. During the reprocessing and pyrolysis of fluoropolymers,
25
the possibility exists for the production of perfluoroisobutene, PFIB, which is a
cause for concern, even in small amounts, because of its extremely high toxicity
(the reported 2 hour lethal dose for PFIB in rats is 1 ppm). It has been found that
o
PTFE begins to form PFIB at 525 C, while other fluoropolymers such as
poly(tetrafluoroethylene-hexafluoropropylene) copolymers form PFIB at tem-
o 25
peratures as low as 380 C. The toxic nature of gaseous products of
fluoropolymers emphasizes the need for good exhaust ventilation during the re-
processing of such polymers.
This problem is exacerbated where PTFE is used as a flame retardant addi-
tive in engineering plastics such as polyphenylene ether, polycarbonate and
26
ABS where it imparts non-dripping characteristics during burning of the poly-
mer. In the temperatures encountered in polymer incineration, PTFE would al-
most certainly produce dangerous levels of PFIB.
CONTAMINATION DURING PROCESSING
METAL CONTAMINATION
Metal contamination (also referred to as “tramp” metal) can processing

equipment itself (such as wear fragments from extruder gates, granulators, alu-
minum adapters, etc.). Small fragments of metal can cause blockage to injection
nozzles of injection molding machines resulting in defective moldings and nozzle
damage. In addition, ferrous/ferric ions due to their multivalent nature, act as
oxidation catalysts through redox reactions. Processing of PVC waste can also
promote acid-induced corrosion of the extruder components resulting in rust
particle contamination.
GEL CONTAMINATION
Recycling of PCR HDPE (usually milk bottles) by melt extrusion can lead to
crosslinking during the thermal reprocessing stage since the antioxidant added
27
initially (during manufacture of the polymer) is consumed. Loosely
crosslinked regions (known as ‘gels’) can acts as stress concentrations in film
and cause ‘blow-outs’ in bottles made from recycled HDPE.
DEGRADED POLYMER CONTAMINATION
A common source of contamination in recycled HDPE (as well as virgin

HDPE) is ‘black specks’. These are small areas of highly degraded polymer that
have been carbonized due to excessive residence time in an extruder. These
‘black specks’ typically occur in low flow regions in the extruder where
‘hang-ups’ form. Often, such contamination may also appear yellow, brown, or
amber depending on the extent of degradation. In fact, under microscopic exam-
ination, a color gradient may be evident indicating that the darkest side has
been in direct contact with the wall of the extruder barrel. Black specks cause a
major problem in the blow molding of natural or white bottles where they are
aesthetically undesirable.
In PET, degraded resin impurities can also arise as a result of reprocessing.
Such impurities are often self-generated during the melt reprocessing step.
During extrusion of PET, oligomers (both linear and cyclic) can form as a result
of polymer break-down. Such oligomeric impurities diffuse towards the surface
of the recycled PET fibers or film and greatly affect their surface properties and
hence adhesion. Problems with printability, dyeability, and treatment can often
22
occur. Oligomeric products of PET include terephthalate esters (e.g.,
monohydroxyethyl-terephthalate, monomethyl-terephthalate and
bis-hydroxyethyl-terephthalate). These oligomeric contaminants can also lead
22
to discoloration (i.e., yellowing) of the recycle.
DETECTION OF CONTAMINATION
A number of patents have been issued in which the nature of impurities
and their detection in recycled plastic is achieved by measuring their IR emis-
28
sion and where the post-consumer plastic itself is categorized by IR imaging. A
novel method for detecting contamination in mixed plastics uses a pulsed laser
beam to form a plasma which is then spectrally analyzed. A rapid technique for
determining the level of PVC in mixed plastic waste has been described where
the commingled polymer powder is thermally degraded using a linear tempera-
o
ture ramp from 260-360 C and the vinyl chloride polymer content is calculated
29
from the HCl concentration in the evolved gas. PVC contamination is of partic-
ular concern in polymer recycling since it can adversely impact upon the recy-
cling process (see Table 3).
Table 3: The effect of PVC contamination on various methods of recycling
Classification Method Effect of contamination

primary in-house scrap recovery polymer discoloration due to conjugated sequences
evolution of HCl at 170oC; machinery corrosion; dis-
secondary melt reprocessing
coloration and black specs in PET
pyrolysis formation of CaCl2 which blocks reactor tubes
tertiary
liquefaction poisoning of cracking catalyst
quaternary incineration HCl and possibly chlorinated dioxins
An automated system developed by Magnetic Separation Systems, Inc. can

segregate commingled plastic containers and remove contamination through
the use of special sensors. First, a mechanical device debales the bottles and
screens out small size and non-plastic contamination. The containers are then
presented to an optical sensor which can identify PET, PP, natural and colored
HDPE. Another sensor is used to further identify green and amber PET from
clear PET containers, while a surface sensing x-ray based sensor identifies PVC
30
containers.
REMOVAL OF CONTAMINATION
REMOVAL OF CONTAMINATION IN POLYOLEFINS
Equipment has been developed for the recycling of HDPE containers,

waste oil and detergent bottles into saleable flakes at rates of 1000 to 5000 lbs
per hour. The most used system is based on hydrocyclone separation, a process
which is effective in separating various types of plastics while virtually elimi-

nating paper and aluminum contaminants. This represents a major improve-
31
ment in quality and cost saving over the “float sink” process.
32
Schimpf describes a process for separating polyolefins from household
and industrial plastic waste which relies on a two step process comprising firstly
a sink-and-float separator and then a hydrocyclone. The hydrocyclone separa-
32
tion allows separation of polystyrene and PVC.
A process to remove polyester and cellulose contamination from PP bale
wrapper for cotton and polyester fibers involves treating the PP with aqueous
sodium hydroxide solution followed by an oxidizing agent such as aqueous so-
33
dium hypochlorite.
REMOVAL OF CONTAMINATION IN PET
A method has been devised for the microseparation of PVC from PET in
which the PVC is subjected to a process of selective bulking which causes it to
34
float. This method is ideally suited for the removal of small quantities of PVC
from large quantities of PET bottle recycle as is generally the case in the USA.
However, in Europe, PVC can be a major component in the recycle stream and
thus must also be recycled. Froth flotation of PVC contaminants provides an ef-
35
ficient way for PVC removal in the recycling of PET bottles.
A process for separating PET and PVC has been recently patented in which
the commingled stream is treated with NaOH and a nonionic surfactant (e.g.,
TM o
Neodol 91-6) in order to reduce the contact angle of PET flakes to less than 25
o
whilst maintaining the contact angle of PVC greater than 45 . Then, the flake
mixture is added to water and agitated to allow PVC flakes to come in contact
with gas bubbles causing the PVC flakes to float and the PET flakes to sink. The
36
system can separate a 50:50 mixture of PET and PVC to yield 97% pure PVC.
Another method for separating PVC from PET relies on air flotation
whereby the PVC is separated through a combination of vibration and flotation,
to ensure separation on the basis of specific weight. This method is especially
1
suited to separating PVC labels from PET bottles.
A recent patent describes a procedure to remove foreign material from
post-consumer PET. The process involves the melt depolymerization of PET by
feeding the melt from an extruder into a special separation vessel where
low-density contamination floats to the surface of the melt and high-density con-
tamination sinks to the bottom of the melt and purified oligomeric PET is re-
37
moved from the intermediate region.
A method for removing the contaminants in the recycling of PET waste has
38
also been devised. In this process, PET is transesterified by heating with ethyl-
ene glycol to give a solution containing short chain PET and bis(2-hydroxyethyl)
terephthalate. By hydrolysis of this mixture at elevated pressure and tempera-
ture, ethylene glycol and terephthalic acid crystals are produced. Activated car-
38
bon and clay are used to absorb contaminants for separation by filtration.
REMOVING CONTAMINATION IN PVC
Contamination of PET in PVC can be successfully removed by melt filtra-

tion through continuous screen changing equipment at a processing tempera-
o
ture of 204 C which is below the melting point of PET and thereby allows PET
3
and other solid contaminants to be filtered out.
Recycled PVC from post-consumer bottles can be purified from contami-
nants other than PET (e.g., paper, aluminum foil, HDPE, PP, polycarbonate,
crosslinked PVC, and fluoropolymers) by melt filtration specially designed to
39,40
maintain low back pressure.
REMOVING CONTAMINATION IN PC
Polycarbonate can be purified during the recycling operation by dissolving

it in an organic solvent. Thus insoluble contamination (e.g., glass fibers, paints,
or metals) can be separated by filtration via special filter beds. Also, unwanted
colorants can be absorbed in this step. A solvent process for recycling PC scrap is
41
currently used.
CONCLUSIONS
Due to the myriad of sources of contamination in polymer recycle, the effect
of contamination is difficult to predict and even more difficult to quantify. The
plastics recycler may not be aware of the presence of some contamination since it
is present in such low amounts. It should be bourn in mind that, despite the pos-
sible low levels, some contamination, for example, transformation products of
flame retardants are extremely toxic and thus can pose a health risk to opera-
tors of the recycling process and end-users of the recycled product. Thus in all
forms of recycling, especially pyrolysis and incineration, a careful approach is
necessary in order to avoid the production of supertoxic compounds.
In other areas, the presence of contamination may not pose any health risk
but simply lowers the quality of the recycle and devalues the economics of the re-
cycling process.
REFERENCES
1. E. Sereni in Recycling of Plastic Materials, Ed. F. P. La Mantia, ChemTec
Publishing,
Toronto, Canada, 1993, p. 99.
2. F. Rodriguez, L. M. Vane, J. J. Schluter, and P. Clark, ACS Symp. Ser., 509, 99 (1992).
3. R. T. Gottesman, Makromol. Chem., Macromol. Symp., Ed. W. Heitz, 57, 115 (1992).
4. R. A. Fleming, Makromol. Chem., Macromol. Symp., Ed. W. Heitz, 57, 75 (1992).
5. Anon. in PP OK in recycled HDPE, Chem. Week, 152, 45 (1993).
6. P. S. Hope, D. A. G. Parsons, G. Capaccio, and M. J. Kitchiner, Makromol. Chem.,
Macromol. Symp., Ed. W. Heitz, 57, 383 (1992).
7. C. Pattanakul, S. Selke, C. Lai, and J. Miltz, J. Appl. Polym. Sci., 43, 2147 (1991).
8. P. S. Blatz in Properties of high density polyethylene from post consumer
recycled containers, Polym. Prep., 32, 152 (1991).
9. R. May in Pesticide Container Recycling, Florida Recycling Fair, May, 1990.
10. S. C. Zahos in Building Momentum in Pesticide Container Recycling, Florida
Recycling Fair, February, 1991.
11. E. H. Neumann in Recycling of Plastic Materials, Ed. F. P. La Mantia, ChemTec
Publishing, Toronto, Canada, 1993, p. 17.
12. J. Goldstein and L. Erwin, Annu. Tech. Conf.- Soc. Plast. Eng., 50, 260 (1992).
13. C. Llop and A. Perez, Makromol. Chem., Macromol. Symp., Ed. W. Heitz, 57, 115
(1992).
14. F. P. La Mantia and M. Vinci, Polym. Degrad. Stab., 45, 121 (1994).
15. J. Troitzsch, Proc. 4th Meet. Fire Retard. Polym., Sept., Freiburg, 1992.
16. H. Meyer, M. Neupert, W. Pump, and B. Willenberg, Kunststoffe, 83, 253 (1993).
17. W. Lorenz and M. Bahadir, Chemosphere, 26, 2221 (1993).
18. G. Camino, M. P. Luda, and L. Costa in Chemical Industry and Environment;
General Aspects- Risk Analysis, Ed. J. Casal, Vol. 1, p. 221, 1993.
19. N. Grassie, Polym, Degrad. Stab., 30, 3 (1990).
20. P. M. Phillips in Analysis of color in recycled resin, ANTEC’93, p. 998, 1993.
21. M. D. Read and K. Mulka in Effect of post-consumer high-density polyethylene
on the color of pigmented resin formulations, ANTEC’92, p. 1295, 1992.
22. W. De Winter in Recycling of Plastic Materials, Ed. F. P. La Mantia, ChemTec
Publishing, Toronto, Canada, 1993, p. 1.
23. F. P. La Mantia and D. Curto in Recycling of Plastic Materials,
Ed. F. P. La Mantia, ChemTec Publishing, Toronto, Canada, 1993, p. 27.
24. M. Bakhramov, A. Abdukadirov, A. Kh. Gafurov, and S. I. Isamukhamedov, Plast.
Massy, 3, 32,48 (1993). Chem. Abs. 120: 78801c, Chem. Abs. 120; 78803e.
25. B. B. Baker and D. J. Kasprzak, Polym. Degrad. Stab., 42, 181 (1993).
26. H. J. Kress, L. Bottenbruch, M. Witman, K. Kircher, C. Lindner, and H. K. Ott,

German Patent DE 3,442,281 (1985); Chem. Abs., 105: 44153t.
27. J. Scheirs, S. W. Bigger, and O. Delatycki, Polymer, 30, 2080 (1989).
28. J. Stor-Pellinen, Y. Pylvanainen, R. Lehtiniemi, J. Rantala, J. Hartikainen and
M. Luukkala, Rev. Gen. Therm., 32, 7 (1993).
29. K. Saito, T. Fukuda, H. Natatani, and K. Kinoshita, Japanese Patent,
JP 04,172,248 (1992).
30. G. R. Kenny, D. R. Morgan, and A. H. Stephens, Proc. Natl. Waste Process. Conf., 15,
307 (1992); Chem. Abs., 120: 166162a.
31. Anon., Plastics World, April 1991, p. 30.
32. A. Schimpf, Kunststoffe, 82, 908 (1992).
33. K. Al-Jumah and L. Tenney, World Patent WO 93 10, 916 (1993); Chem. Abs., 120:
166765x.
34. H. Frankel, Proc. of Recycle 92, Davos, Switzerland, April 7-10, 1992.
35. J. Maczko and R. Kobler, Polym. Prep., ACS Div. Polym. Chem., 34, 920 (1993).
36. E. A. Sisson, Eur. Pat. EP 512,464 (1992).
37. E. J. Burkett and R. S. Jenks, US Patent 5,223,544 (1993).
38. S. M. West, International Patent WO 93 23,465 (1993); Chem. Abs. 120: 219756f.
39. G. M. Sadowski and L. G. Shaw, J. Vinyl Technol., 14, 135 (1992).
40. H. K. Boo, B. E. Mikofalvy, J. W. Summers, W. A. Sell, and D. H. Mittendorf, J. Vinyl
Technol., 14, 140 (1992).
41. B. Miller in How to blow mold recycled resin, Plast. World, Dec. 1990, p. 21.
G. Wypych 185
UV Stability of Recycled Polymers
George Wypych
ChemTec Laboratories, 38 Earswick Dr., Toronto, Ontario MIE 1C6, Canada
Several decades of UV degradation studies gave sufficient evidence to

make the following statements:
• all polymers degrade when exposed to UV light
• products of thermal degradation typically accelerate UV degradation.1
From these two statements, it is obvious that strategy of recycling must
consider material durability as very important factor.
Recycling would still be a relatively simple process if materials for recy-
cling were identified and composed of separated, individual polymers but they
are heterogeneous mixtures of unknown composition.
The practical reality causes that recycled materials are most likely poly-
mer blends either due to the necessity or because of technological choice. Deal-
ing with two or more components further complicates UV stability because
polymers in mixtures interact.
The above complex situation suggests that without a good plan, UV stabil-
ity problems will never be resolved for polymer mixtures, which is already con-
firmed by very little understanding of principles governing UV stability of
polymer blends today. Literature on this subject is very limited and was already
2,3
reviewed elsewhere. Below, only such examples from literature are reviewed
which may contribute to the formulation of principles leading to a design of sta-
ble polymer blends for further application in recycling technology.
186 UV stability of recycled polymers
Table 1: Polymer type classification based on photochemical behavior [after J.

Lemaire, Report PB90-2002482, NTIS, 1990]
Type Description Typical examples

polymer non-absorbing; photooxidation depends on normal ABS, EMAc, EPR, NBR, PB, PBT, PEBA,
A
structure alone PET, PI (isoprene), PMP, PU (ether), SBR
polymer non-absorbing; photooxldation depends on normal

B
structure and no-absorbing chromophores EEA, EVA, LDPE, LLDPE, PP
polymer absorbing; direct photochemical reaction of absorbing

C groups and photooxidation controlled by chemical defects which PA-6, PA-11, PA-12, PC, PS, PU (ester), SAN
are not chromophores
polymer non-absorbing; photochemistry controlled by chemical

D
defects and chromophore impurities PVC, PVDC, PVDF, SI
Table 1 gives polymer classification based on absorption and utilization of

4
radiant energy by polymers. Group A includes several polymers which are more
UV stable than polymers included in the other groups and the reasons are perti-
nent from description given in the second column of the table. Polymers included
in group C are more likely to be degraded by UV radiation because of their inher-
ent weakness - their structural components absorb radiation. PMMA and PC
can serve as examples of representatives of these two groups, and known details
of their stability fit principles of classification. If these two polymers are mixed,
the resultant blend is more stable than PMMA.
5,6
Papers by Osawa et al. include general suggestions of choice of polymer
pair. The PC/PMMA blend was studied under varying radiation wavelength. In
the region where both polymers can absorb, PC, having stronger absorption,
shields PMMA from radiation and even small additions of PC show stabilizing
effect on PMMA. Presence of PMMA does not have any effect on PC. When blend
is exposed to radiation above 290 nm (not absorbed by PMMA) only PC degrades
with a relatively low rate and fact of inclusion of PMMA in the blend have no
bearing on PC degradation rate. Two principles are involved in this experiment:
• if polymer does not absorb radiation it cannot decrease the stability of the
other polymer (it may only participate in quenching of radicals formed by
the second polymer in blend - stabilization)
• if polymers reside in separate phases their interaction is limited to their in-
terface.
G. Wypych 187
Let us analyze further reasons explaining these observations. Table 2 gives

a list of initiators of degradation and products of degradation for some common
1
polymers.
It is noticeable that PMMA degradation is accelerated only by a few admix-
tures, and perhaps it is an additional reason for a well-known durability of this
polymer. Products of PC degradation can only contribute to PMMA degradation

if there is an immediate contact between the two polymers which only happens
for a small fraction of the material.
Table 3 is composed based on technical data for over 180 commercial
7
blends. Producers of blends included in Table 3, rated their products as weather
resistant materials for reasons included in the last column. Analysis of this in-
formation shows that most blends included in the table have at least one poly-
mer which belongs to group A defined in Table 1. It means that weatherstable
blend is more likely to be obtained from combination of durable material. In fact,
a more thorough analysis of information in Table 3 shows that UV stability of a
mixture is at best an average of composition. Only in two cases, blend stability
was better than expected from component polymers. At the same time, litera-
ture on the subject contains numerous examples of compostions which are less
2
durable than component polymers.
From the data in Table 2, it is very easy to predict that, for example, blend
of PA and PEST is not stable because PA produces water, amines, and acids as
degradation products, and they are accelerators of PEST degradation (perhaps,
it is one of the reasons that such blends are not common). Similarly, PVC is not
the best candidate for mixing with PC or PEST because HCl evolved can acceler-
ate their hydrolysis. It should be noticed from Table 2 that additives play also
very essential role in degradative processes. For example, mixing of polymer
pigmented with TiO2 with PA or PE enhances their degradation.
Several other examples can be given to formulate principles of formation of
stable polymer compositions. Figure 1 shows that when ethylene-propylene-
rubber, EPR, concentration in the blend with high density polyethylene, HDPE,
8
is increased above 30%, the UV stability of blend rapidly increases. Detailed in-
vestigation of reasons for this sudden change shows that around this concentra-
tion (30-40% EPR) surface becomes composed of EPR. Further increase in EPR
concentration only increases the thickness of the skin but blend stability re-
mains the same because it is controlled by EPR chemistry throughout this
range. Such use of blend incompatibility suggests fundamental alternative to
chemical durability, especially, when one considers that the durability of sur-
face determines material stability. UV light has only limited penetration (up to
0.15 mm depending on composition, most frequently less than 0.01 mm).
G. Wypych 189
Figure 1. Number of cycles vs. EPR concentration in HDPE/EPR blend. [Data from S. A. Komarov,
E. E. Piskunova, V. N. Kuleznev, V. V. Korovkin, A. A. Kolesnikov, and I. K. Prik, Kolloidn. Zh., 48,
152, (1986)
The same effect can be achieved by two other methods: technological and
physical. Frequent attempts are made recently to coextrude materials (paper by
Uhlen in this book is one of the existing examples). The surface is composed of
virgin material to protect recycled material which is coextruded as an internal
layer. The physical method involves similar principle as given in the case of
9
EPR/HDPE blend, but materials involved have very different surface energies.
G. Wypych 191
Interesting paper, which belongs to this group, was published by Arnold et

9
al. on a blend of poly(siloxane imide) segmented copolymers with
polybenzimidazole intended for spacecraft application as coating or adhesive. It
is interesting to note that the poly(siloxane imide) due to its low surface energy
was expected to dominate surface. At the same time materials for this applica-
t ion req uir e r es is ta nc e t o a t o m i c o x y g e n a t m o sph e re t o w h i ch
polybenzimidazole is not resistant. Intention of this combination was that copol-
ymer found on surface will decompose to silicate layer which may form protec-
tive layer. XPS study and atomic oxygen stability tests confirmed that such was
the mechanism of action of this blend. This example is a part of a large effort in
which small additions of various polysiloxanes are used to form microlayers on
the surface of various polymers. The technology used here is termed as surface
enrichment which can exceed 99% of polysiloxane in the surface layers. A few
seconds of treatment of material by plasma or UV radiation converts organic
material to an impenetrable layer of SiO2 which not only gives UV protection but
also changes various physical characteristics of material such as gas permeabil-
ity, abrasion and chemical resistance, etc.
It seems reasonable to expect that the stability of blends depends on their
10
morphology. It was stressed that studies of polymer degradation mostly focus
on chemical mechanisms whereas polymer morphology is an important contrib-
utor to its overall stability. This point can be very well illustrated by data of
11
Chien at al. (Figure 2). Polystyrene and poly(vinyl methyl ether) form miscible
blends. If these blends are exposed to UV degradation molecular weight rapidly
changes for both polymers. Only when PS/PVME ratio is 70/30, the molecular
weight changes slow down. Blends containing less than 70% PS become cloudy
on exposure and display two Tgs (compared to one Tg for initial blend). These ob-
servations indicate that oxidation causes phase separation of the blend. When
both polymers are in good contact, PS undergoes changes and PVME changes
are slowed down. These observations show that products of degradation of one
polymer cross-propagate changes in another polymer unless two polymers in
mixture are phase separated.
It should be noted that reverse is true if blend is stabilized by chemical re-
12
action. Such case was demonstrated for PVC/PBR blend. In the reported case
12
of PVC/PBR blend, the addition of PBR to PVC was only effective up to a cer-
tain concentration which coincided with limiting concentration of miscible
Figure 2. Effect of 26 days aging on molecular weights of PS and PVME. [Adapted, by permission,
from Y. Y. Chien, E. M. Pearce, and T. K. Kwei, Polym. Prepr., 29, 2086 (1988).
blend formation. When phase separation is apparent the probability of chemical

reaction, stabilizing blend, decreases as does the blend overall stability.
A proper distinction must be made, in this respect, between radicals and
low molecular products of reaction. Radicals are short-lived therefore the bar-
rier of different phases is sufficient to limit cross-propagation. Low molecular
products of degradation move within the material very easily and their effect
(catalysis of degradation, chemical reaction, etc.) cannot be restricted by phase
separation.
The analysis of the above examples shows that present choices for forma-
tion of stable polymer blends are very limited, and can be well described by the
following points:
• Practical implication of plastic recycling in form of polymer mixtures de-
lays and aggravates problem of waste accumulation rather than solves it
• Many polymers have antagonistic properties in respect to UV stability
which shortens their effective life. In most cases, processing of combination
of polymers which increase each others degradation rates will not allow to
use these mixtures in future for any useful recycling purposes
G. Wypych 193
• The best combination of two polymers recycled together includes at least

one polymer having high durability
• Polymers and their additives should be paired in such a manner as to ex-
clude their cross-catalysis and cross-propagation
• In most cases polymers residing in separate phases give the best opportu-
nity to increase span-life of polymer beyond single recycling. This implies
that recycled polymers should be used as fillers, in phase-separated blends
which can be simply achieved by processing polymers differing in melting
point
• Since most degradation occurs on surface, the surface of materials should
be preserved by one of the existing methods (screening, surface enrich-
ment, etc.) if plastics are to be effectively recycled
• Methods should be developed to remove products of degradation from poly-
mer bulk
• Development of compatibilizers is required which can isolate polymer do-
mains (prevent cross-propagation) or react with products of degradation
• Considering future recycling more emphasis should be given to production
of UV stable polymers and streamlining production in such a manner that
it will make collection of secondary materials suitable for recycling bearing
in mind that complex polymer mixtures make recycling process more diffi-
cult.
REFERENCES
1. G. Wypych in Weathering Handbook, ChemTec Publishing, Toronto, 1990.
2. G. Wypych, Polym. Networks Blends, 2, 53 (1992).
3. F. P. La Mantia in Handbook of Polymer Degradation, Eds. S. M. Hamid and
M. B. Amin, M. Dekker, New York, 1992.
4. J. Lemaire in Long-term Forecasting of the Behavior of Polymer Materials in
Outdoor Weathering Conditions, Report PB90-2002482, NTIS, 1989.
5. Z. Osawa and Y. Fukuda, Polym. Degrad. Stab., 32, 285 (1991).
6. Y. Fukuda and Z. Osawa, Polym. Degrad. Stab., 34, 75 (1991).
7. L. A. Utracki, Ed., Encyclopaedic Dictionary of Commercial Polymer Blends,
ChemTec Publishing, Toronto, 1994.
8. S. A. Komarov, E. E. Piskunova, V. N. Kuleznev, V. V. Korovkin, A. A. Kolesnikov, and
I. K. Prik, Kolloidn. Zh., 48, 152, (1986).
9. C. A. Arnold, D. H. Chen, Y. P. Chen, R. 0. Waldbauer, M. E. Rogers, and
J. E. McGrath, High Perform. Polym., 2, 83 (1990).
10. G. Wypych in Poly(vinyl chloride) Degradation, Elsevier, Amsterdam, 1985.
11. Y. Y. Chien, E. M. Pearce, and T. K. Kwei, Polym. Prepr., 29, 2086 (1988).
12. M. K. Naqvi and A. R. Sen, Polym. Degrad. Stab., 33, 367, (1991).
Index 195
INDEX
A contaminants 168
acrylics 135 contamination 118, 167
additives 173 by other polymers 169
adhesion 77, 82, 162 effect 179
alloys 147 metal 177
aluminum 56 non-polymeric 171
automatization 118 removal 179
automotive 169 copolymer
block 78
B grafted 129
bitumens 129, 130, 157 triblock 79
blends 70, 77, 129, 147 copolymers
stability 188 grafted 130, 147
blow molding 66 copolymers synthesis 131
bottles 51, 54, 65, 168, 171 cross-propagation 192
branching 69 crystallinity 65, 66
building products 169 cups 168
C D
CaO 93 decrosslinking 130
capillary viscometer 79 degasification 118
cellulose 170 degradation
chain branching 65 chemical 133
coextrusion 51, 57, 129 initiators 187
compatibility 77 mechanical 132
compatibilization 77, 78, 86 photooxidative 66
mechanism 85 PVC 94
compatibilizer 72, 74, 78, 130, 148, ultrasonic 133
150, 152 dehydrochlorination 26, 52
compatibilizers 93, 96, 107, 111, 112 depolymerization 130
compatibilizing 129, 130 diffusion 172
composition 44, 125, 126 dismantling 44, 45
compression testing 160 dissociation
196 Index
constants 141 flotation 78

dissociation 139 flow rate 106, 113
drying 47, 118 flow stability 60
foaming 59
E furniture 169
elastic modulus 99, 102, 107, 110,
113 G
elongation 65, 66, 71, 73, 75, 88, GC-MS 172
100, 105, 108, 111, 126 GPC 31, 144, 145
energy recovery 24 grafting 142, 144
EPDM 106, 115 greenhouse film 65
EPR 188 grinding 56
EVA 96, 162
EVA/PS 157 H
EVAc 96 HCl 24, 34
extruder 79 HDPE 125, 168, 169, 170, 172, 188
shear rate 82 homogenization
extrusion 94 silos 118
conditions 79, 97 homomicronization 94, 117, 118
pressure 68 chamber 118
repeated 65, 68 equipment 122
shear stress 86 hopper 124
temperature profile 61 hydrocyclone 121
torque 85, 86 hydrogenation 52
F I
ferrous materials 118 image analysis 79
fillers 33, 102, 103, 104, 106 impact resistance 59
inert 93, 96 impact strength 80, 88, 89, 90, 100,
polymeric 94, 130 103, 104, 105, 108, 111, 115
film incineration 24
mulch 173 incinerators 52
x-ray 176 incompatibility 73
filtering 118 initiator 130
fire retardant 175 injection molding 79, 177
flexural modulus 161 IR 30, 118
Index 197
IR imaging 179 oxygenated compounds 65

irradiation 133 ozone 129
ozonization 129, 134
L mechanism 136
landfill 23, 43, 52
LDPE 71, 167, 176 P
LDPE/PA-6 71 PA 71, 188
lubrication 59 packaging 171
lubricant 95, 98, 176 particle
diameter 84
M distribution 83, 84
macroinitiators 134 size 45, 82
macromonomers 135 PC 181
macroradicals 132 PC/PMMA 186
maleic anhydride 78, 130 PE 56, 70, 78, 93, 98, 132, 135, 145,
material shape 64 146, 175
mechanical properties 36, 70, 77, PE/PS 149, 153
88, 93, 97, 101, 125, 149 PE/PVC 148, 150
melting point 64, 80, 81, 94, 117 PE/PVC/PS 155
milk 172 pelletizing 118
miscibility 78 PEST 188
modification 176 pesticides 172
modifiers 93, 96, 105 PET 56, 77, 78, 93, 167, 169, 170,
moisture 173 172, 173, 176, 180
molecular weight 31, 64, 66, 67, 159 PET/HDPE 107
morphology 66, 67, 82, 86 PET/PP 80, 84, 87
phase
N adhesion 77
Newplast process 118 crystalline 80
NMR 144, 145 dispersion 147
photodegradants 175
O PIB 79, 132
organochlorine compounds 172 pigments 175
oxidation 134 pilot plant 55
oxidation catalyst 178 pipes 51, 52, 53
198 Index
plastic waste R
containers 93 radical polymerization 135
heterogeneous 63, 93 recombination 132, 146
homogeneous 63 recycled polymers
industrial 118, 171 characterization 125
mixed 117 residual stability 26, 39
urban 78, 118 rheological properties 78, 85
PMMA 143, 187 roofing sheets 29
polarity 78 rubber
polyaddition 136 functionalized 77, 83
polydispersity 159
polymer S
classification 186 SBS 93
degradation 64, 117 SEBS 96, 115
incompatibility 64 SEM 72, 74, 79, 81, 109, 112
polymers separation 56, 78, 118, 180
crosslinked 129 color 45
multilayer 170 metal 47
polyolefins 77 paper 118
PP 56, 68, 73, 77, 78, 135, 146, 170 rubber 46
processability 86 wood 46
processing separator
conditions 65 friction 121
PS 31, 125, 191 high density 120
PS/PVC 154 magnetic 120
PS/PVME 191 shredding 118
PTFE 177 soil 173
PU 175 sorting 47
PVC 23, 27, 43, 51, 93, 98, 117, 135, stability
145, 169, 179, 180, 181 thermal 57
determination 97 UV 185
PVC/PBR 191 stabilization 32
PVDF 143, 146 stabilizers 44, 98
PVDF/PS 157 heat 28
PVME 191 lead 30
pyrolysis 52 metals salts 30
Index 199
thermal 95
UV 175 U
stress-strain curve 89 unsaturations 65
structure 67
surface enrichment 191 V
Vicat temperature 90
T viscosity 68, 87
tensile strength 71, 77, 88, 99, 102, voids 77
104, 107, 110, 114, 126, 148, 150
Tg 81 W
thermogram 80 washing 118
thermoplastics 78 windows 30, 32, 43
titanates 93, 97
titanium dioxide 188 X
titration 138 XPS 191
torque 144
transition Y
brittle-ductile 88 yield stress 126
transportation
pneumatic 118

Recycling of PVC and Mixed Plastics Wastes by F. P. La Mantia

Uploaded by

Copyright:

Available Formats

You might also like

Recycling of PVC and Mixed Plastics Wastes by F. P. La Mantia

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Recycling of PVC and Mixed Plastics Wastes by F. P. La Mantia

Uploaded by

Copyright:

Available Formats

Francesco Paolo La Mantia, Editor

Canadian Cataloguing in Publication Data

La Mantia, Francesco Paolo

Includes bibliographical references and index.

1. Polyvinyl chloride--Recycling. 2. Plastic scrap--

TP1122.L35 1996 668.4’236 C95-920960-3

Investigations on the Recycling Ability of Used PVC 23

Determination of Molecular Weight 31

Recycling of Complete PVC Windows 43

Recycling PVC Bottles and Pipes by Coextrusion 51

Basic Concepts on the Recycling of Homogeneous and

Compatibilization of Recycled Polyethyleneterephthalate/

Recycling of Post-consumer Plastic Containers for Liquids 93

A New Process for Recycling of Mixed Plastics Waste 117

Grafted Copolymers and Their Use in Polymer Blends 129

2.3 Transfer to Polymer 134

Effect of Contamination on the Recycling of Polymers 167

UV Stability of Recycled Polymers 185

1 POLYMERS IN WASTE MANAGEMENT

The technology of composting to give added value to domestic waste is likely to

1.3 MATERIALS RECYCLING OF SEGREGATED, CLEAN INDUSTRIAL WASTE

Some polymers, (e.g. polymethylmethacrylate, polystyrene) lead to good yields

Table 1: Composition of gaseous products from the pyrolysis

1.5 INCINERATION WITH HEAT RECOVERY

Disposing of plastics by burial in sanitary landfill, contrary to popular belief, is

Scheme 1: Mechanooxidation of PVC18

Scheme 2: Sensitisation of PVC degradation by hydroperoxides

concentrated on eliminating the unsaturated components of the impact modifi-

4.3. “COMPATIBILISER” (SOLID PHASE DISPERSANTS)

PVC often appears as a relatively minor component in mixed plastics waste in

Since HCl is such a potent redox initiator for the decomposition of

5.2 REMOVAL OF CONJUGATED UNSATURATION

By the incorporation of a maleate entity in a tin compound (I) complemen-

ROCOCH2SH + CH2=CH- → ROCOCH2SCH2CH2- [6]

Reaction 6, where R contains a hindered phenol or a UV absorber function,

Compounds containing sulphide groups are widely used as synergistic antioxi-

Scheme 3: Photosensitizing action of the tin thioglycollates

5.5 CHAIN-BREAKING ANTIOXIDANTS

By themselves, chain-breaking donor antioxidants such as the hindered phe-

5.6 REAGENTS FOR ALLYLIC CHLORINE

The concentration of V is an order of magnitude higher than that of VI, but

Allylic methylene is much more strongly implicated under the oxidative

6 ANTAGONISM AND SYNERGISM

In general, an effective light screening pigment such as carbon black or ti-

Table 2: Sulphur-containing antioxidants as photoantioxidants

An alternative approach to stabiliser substantivity under aggressive con-

Scheme 4: Antioxidant adduct formation in PVC during processing

Scheme 5: Maleic ester adduct formation in PVC

able. In particular, degradability and composting are not suitable because

D. Gilead in Degradable Polymers, Principles and Applications, Eds G. Scott

D. Braun and K. Krämer

Deutsches Kunststoff-Institut, Schloßgartenstraße 6, 64289 Darmstadt, Germany

The separation in a hydrocyclone, which works based on the principle of sorting

The main disadvantage of PVC is the rather limited thermal stability

Figure 2. Schematic formula of the dehydrochlorination of PVC.

he main function of heat stabilizers is to prevent degradation during pro-

The PVC-sample (≈ 0.1 g) is introduced into the degradation vessel and

nation is performed by continuous conductometric measurements. As a result

Figure 4. Schematic degradation curve of stabilized PVC.