Recycling packaging materials R. Franz and f. welle, fraunhofer Institute for process engineering and Packaging, Germany 23.1 Introduction Food packaging is a still growing market. As a consequence, the demand to re-use post-consumer packaging materials grows as well. Recycling of packaging materials plays an increasing role in packaging, and numerous applications can already be found on the market. Ten or twenty years ago most post-consumer packaging waste was going into landfill sites or to incineration. Traditionally, only glass and paper/board were recycled into new applications. In the case of packaging plastics the situation is quite different. Only uncontaminated in-house production waste was collected, ground and recycled into the feedstream of the packaging production line without further decontamination. With increasing environmental demands, however, post-consumer plastics packaging materials have also been considered more and more for recycling into new packaging A closed-loop recycling for packaging plastics is also supported by public pressure. The packaging and filling companies have to take responsibility for their packaging materials and environmental concerns. In many countries the consumer, government and the packaging companies want to have packaging materials with a more favourable ecobalance in the supermarkets. A more favourable ecobalance can be achieved with different approaches. One of these approaches is the re-use of recycled material in packaging. This development is driven by the recent strong increase in polyethylene terephthalate(PET) bottles used for soft drinks, water and other foodstuffs. Today, many filling companies have decided to start using recycled plastics into their PET bottles in the near future But recycling of packing plastics is also a question of recycling technology and collection of packaging waste. Today many countries have established
23.1 Introduction Food packaging is a still growing market. As a consequence, the demand to re-use post-consumer packaging materials grows as well. Recycling of packaging materials plays an increasing role in packaging, and numerous applications can already be found on the market. Ten or twenty years ago most post-consumer packaging waste was going into landfill sites or to incineration. Traditionally, only glass and paper/board were recycled into new applications. In the case of packaging plastics the situation is quite different. Only uncontaminated in-house production waste was collected, ground and recycled into the feedstream of the packaging production line without further decontamination. With increasing environmental demands, however, post-consumer plastics packaging materials have also been considered more and more for recycling into new packaging. A closed-loop recycling for packaging plastics is also supported by public pressure. The packaging and filling companies have to take responsibility for their packaging materials and environmental concerns. In many countries the consumer, government and the packaging companies want to have packaging materials with a more favourable ecobalance in the supermarkets. A more favourable ecobalance can be achieved with different approaches. One of these approaches is the re-use of recycled material in packaging. This development is driven by the recent strong increase in polyethylene terephthalate (PET) bottles used for soft drinks, water and other foodstuffs.1 Today, many filling companies have decided to start using recycled plastics into their PET bottles in the near future. But recycling of packing plastics is also a question of recycling technology and collection of packaging waste. Today many countries have established 23 Recycling packaging materials R. Franz and F. Welle, Fraunhofer Institute for Process Engineering and Packaging, Germany
498 Novel food packaging techniques collection systems for post-consumer packaging waste, like the green dot systems. Such country-wide collecting systems guarantee increasing recovery rates. Together with new developments of recycling systems and with increasing recycling capacity the way is open for some plastics for a high value recycling of packaging waste. Due to health concerns most of the recycled post-consumer plastics are going into less critical non-food applications, but in recent years there have also been efforts to recycle post-consumer plastics like PET into new food packaging applications. This changes the situation for some packaging plastics from an open-loop recycling of packaging plastics into a closed-loop consumer plastics into direct food contact application needs much more knowledge about contamination and migration than for non-food applications, order to assess the risk to consumers' health. Additionally a quality assurance system for post-consumer plastics should be established 23.2 The recyclability of packaging plastics It is generally known that food contact materials are not completely inert and can interact with the filled product. In particular, interactions between packaging plastics and organic chemicals deserve the highest interest in this context. Such interactions start with the time point of filling and continue during the regular usage phase of a package and even longer, in case a consumer misuses' the empty packaging by filling it with chemical formulations such as household cleaners, pesticide solutions, mineral oil or others. The extent of these interactions depends on the sorption properties and the diffusion behaviour which is specific to certain polymer types or individual plastics. These physical properties together with the contact conditions ultimately determine the potential risk of food contamination from recycled packaging plastics. In other words, taking only the polymer itself into consideration and not possible recycling technologies with their special cleaning efficiencies, etc, under given conditions the inertness of the polymer is the basic parameter which determines the possibility for closed-loop recycling of packaging plastics. The inertness of common packaging polymers decreases in the following sequence Poly(ethylene naphthalate)(PEN), poly(ethylene terephthalate)(PET), rigid poly (vinyl chloride)(Pvc)> polystyrene(PS)> high density polyethylene(HDPE), polypropylene(PP)> low density polyethylene In relation to this aspect, PEN, PET or rigid Pvc do possess much more favourable material properties in comparison to other packaging plastics, such as polyolefins or polystyrene and are, therefore, from a migration related point of view much better suited for being reused in packaging applications. Polymers
collection systems for post-consumer packaging waste, like the green dot systems. Such country-wide collecting systems guarantee increasing recovery rates. Together with new developments of recycling systems and with increasing recycling capacity the way is open for some plastics for a high value recycling of packaging waste. Due to health concerns most of the recycled post-consumer plastics are going into less critical non-food applications, but in recent years there have also been efforts to recycle post-consumer plastics like PET into new food packaging applications. This changes the situation for some packaging plastics from an open-loop recycling of packaging plastics into a closed-loop recycling into new packaging materials. However, the recycling of postconsumer plastics into direct food contact application needs much more knowledge about contamination and migration than for non-food applications, in order to assess the risk to consumers’ health. Additionally a quality assurance system for post-consumer plastics should be established. 23.2 The recyclability of packaging plastics It is generally known that food contact materials are not completely inert and can interact with the filled product. 2 In particular, interactions between packaging plastics and organic chemicals deserve the highest interest in this context. Such interactions start with the time point of filling and continue during the regular usage phase of a package and even longer, in case a consumer ‘misuses’ the empty packaging by filling it with chemical formulations such as household cleaners, pesticide solutions, mineral oil or others. The extent of these interactions depends on the sorption properties and the diffusion behaviour which is specific to certain polymer types or individual plastics. These physical properties together with the contact conditions ultimately determine the potential risk of food contamination from recycled packaging plastics. In other words, taking only the polymer itself into consideration and not possible recycling technologies with their special cleaning efficiencies, etc., under given conditions the inertness of the polymer is the basic parameter which determines the possibility for closed-loop recycling of packaging plastics. The inertness of common packaging polymers decreases in the following sequence: Poly(ethylene naphthalate) (PEN), poly(ethylene terephthalate) (PET), rigid poly(vinyl chloride) (PVC) > polystyrene (PS) > high density polyethylene (HDPE), polypropylene (PP) > low density polyethylene (LDPE) In relation to this aspect, PEN, PET or rigid PVC do possess much more favourable material properties in comparison to other packaging plastics, such as polyolefins or polystyrene and are, therefore, from a migration related point of view much better suited for being reused in packaging applications. Polymers 498 Novel food packaging techniques
Recycling packaging materials 499 like polystyrene and HDPE may also be introduced into closed loop recycling if the cleaning efficiency of the recycling process is high enough regarding the input concentrations of post-consumer substances. However, regarding consumers' safety, the composition and concentration of typical substances in post-consumer plastics and the ability of the applied recycling process to remove all post-consumer substances to concentrations similar to virgin materials is of interest. The incoming concentration of post-consumer contaminants can be controlled off-line with laboratory equipment like gas chromatography or HPLC or online with detecting or sniffing devices. With help of online devices nearly a 100% control of the input materials can be established. Therefore the post consumer material is much more under control and packaging materials with high concentrations of migratable substances, or misused bottles, can be rejected and the requirements on the cleaning efficiency of the recycling process are lower. The source control is therefore the crucial point regarding of the worst casescenario of the so-called challenge test (see Section 23. 4.1) Recovery of packaging plastics into new packaging applications requires blending of recycled with virgin materials. In praxis today, the recyclate content of packaging materials varies from only a few per cent up to 50% recycled material in some packaging applications. Numerous studies have been carried out on the determination the material properties and the blending behaviour of recycled plastics. However, it is not the focus of this chapter to deal with blending of polymers but it needs to be stressed that the recycled material should be suitable for blending with virgin materials. Additionally, the mechanical properties of the recyclate should be not influenced in a negative way, so as to avoid potential consequences for the additive status of the recycled plastics The average number of cycles is a function of the blend ratio and the number of recycling steps carried out. In practice the average number of cycles ranges from one to three. Therefore, the material is not recycled many times and the problem of accumulation of degradation products is in most cases of no concern An inherent problem of recycling, however, is the inhomogeneity of the recovered materials. Normally various polymer additives, lubricants, etc. are used by the different polymer manufacturers or converters in order to establish the desired properties of the packaging materials, and all different polymer additives are found as a mixture in the recyclate containing packages. Modern sorting technologies are able to provide input materials for recycling which are nearly 100% of one polymer type. Taking, in addition, the additive status into account will be a sophisticated challenge of future developments. Together with the inertness of the polymers this is one reason why recent closed-loop recycling efforts are focused on polymers which have low amounts of additives e.g. PET However, as mentioned above, the question of recyclability is mainly influenced by the source control of the input material going into the recycling process. If the recovery system considers the manufacturer or the origin of the packaging materials, usually the additive status of the input feedstock is known. An example for this will be HDPE milk bottles collected by a deposit system(see Section 23.5.2)
like polystyrene and HDPE may also be introduced into closed loop recycling if the cleaning efficiency of the recycling process is high enough regarding the input concentrations of post-consumer substances. However, regarding consumers’ safety, the composition and concentration of typical substances in post-consumer plastics and the ability of the applied recycling process to remove all post-consumer substances to concentrations similar to virgin materials is of interest. The incoming concentration of post-consumer contaminants can be controlled off-line with laboratory equipment like gas chromatography or HPLC or online with detecting or sniffing devices. With help of online devices nearly a 100% control of the input materials can be established. Therefore the postconsumer material is much more under control and packaging materials with high concentrations of migratable substances, or misused bottles, can be rejected and the requirements on the cleaning efficiency of the recycling process are lower. The source control is therefore the crucial point regarding of the ‘worstcase’ scenario of the so-called challenge test (see Section 23.4.1). Recovery of packaging plastics into new packaging applications requires blending of recycled with virgin materials. In praxis today, the recyclate content of packaging materials varies from only a few per cent up to 50% recycled material in some packaging applications. Numerous studies have been carried out on the determination the material properties and the blending behaviour of recycled plastics. However, it is not the focus of this chapter to deal with blending of polymers but it needs to be stressed that the recycled material should be suitable for blending with virgin materials. Additionally, the mechanical properties of the recyclate should be not influenced in a negative way, so as to avoid potential consequences for the additive status of the recycled plastics. The average number of cycles is a function of the blend ratio and the number of recycling steps carried out. In practice the average number of cycles ranges from one to three.3 Therefore, the material is not recycled many times and the problem of accumulation of degradation products is in most cases of no concern. An inherent problem of recycling, however, is the inhomogeneity of the recovered materials. Normally various polymer additives, lubricants, etc. are used by the different polymer manufacturers or converters in order to establish the desired properties of the packaging materials, and all different polymer additives are found as a mixture in the recyclate containing packages. Modern sorting technologies are able to provide input materials for recycling which are nearly 100% of one polymer type. Taking, in addition, the additive status into account will be a sophisticated challenge of future developments. Together with the inertness of the polymers this is one reason why recent closed-loop recycling efforts are focused on polymers which have low amounts of additives e.g. PET. However, as mentioned above, the question of recyclability is mainly influenced by the source control of the input material going into the recycling process. If the recovery system considers the manufacturer or the origin of the packaging materials, usually the additive status of the input feedstock is known. An example for this will be HDPE milk bottles collected by a deposit system (see Section 23.5.2). Recycling packaging materials 499
00 Novel food packaging techniques 23.3 Improving the recyclability of plastic packaging 23.3.1 Souree control The source control is the first and most important step of packaging plastics. There must be efficient recovery or sorting processes which are able to control the input fraction going into a closed-loop recycling process. The feedstream material should have a minimum polymer type purity of 99%. Other polymers, which may interfere, have to be sorted out of the recycling stream. Also the first life of the packaging material is of interest. In general only packages previously filled with foodstuffs should be used as an input fraction for a closed-loop recycling process. However there are exceptions, e.g. for Pet due to its high inertness the first packaging application is not so important. Two studies were undertaken",to determine the impact of PET materials formerly used for non-food applications. Both studies came to the same result, that due to the low diffusivity of PET packages from non-food applications could also be used as input material for bottle-to-bottle recycling This underlines the favourite position of PeT bottles for a closed loop recycling It could be shown that deposit systems and recovery systems like curbside ackaging collections with efficient sorting processes, are able to support input materials for high value recycling. However, as mentioned above, the higher the diffusivity of the polymer and, therefore, higher sorption of post-consumer substances the more important is the source control in order to reduce contamination with post-consumer substances or misused packages. The source control can be supplied by modern detecting or sniffing devices which are able to reduce the intake of undesired post-consumer substances into the recycling 3.3.2 Contamination levels and frequency of misuse of recycled plastics Regarding the typical contamination of post-consumer plastics most published data are available for PET bottles and corresponding recyclates. Most of them have quantified or identified substances in post-consumer PET by using different methods. Sadler et al. b, published two studies containing data of contaminants in recycled PET. In the first study he pointed out that most compounds found in recycled PET come from PET starting materials oligomers, flavour bases, label materials and compounds originating in base cups. Contaminants which do not fall into one of these categories are rare. In samples with high levels of contaminants the sum of all compounds was detected to be approximately 25 ppm. No single contaminant appears to be present in post-consumer PET above I ppm and all non-usual compounds in post-consumer PET were present below 0. 1 ppm. In a second study the identity and origin of contaminants in food grade virgin and commercially washed post consumer PET flakes were determined. A total of 18 samples of post-consumer recycled PET flakes was examined. In most cases, positive identification was possible, however, in few cases ambiguity resulted from the similarities in mass
23.3 Improving the recyclability of plastic packaging 23.3.1 Source control The source control is the first and most important step in closed-loop recycling of packaging plastics. There must be efficient recovery or sorting processes which are able to control the input fraction going into a closed-loop recycling process. The feedstream material should have a minimum polymer type purity of 99%. Other polymers, which may interfere, have to be sorted out of the recycling stream. Also the first life of the packaging material is of interest. In general only packages previously filled with foodstuffs should be used as an input fraction for a closed-loop recycling process. However there are exceptions, e.g. for PET due to its high inertness the first packaging application is not so important. Two studies were undertaken4,5 to determine the impact of PET materials formerly used for non-food applications. Both studies came to the same result, that due to the low diffusivity of PET packages from non-food applications could also be used as input material for bottle-to-bottle recycling. This underlines the favourite position of PET bottles for a closed loop recycling. It could be shown that deposit systems and recovery systems like curbside packaging collections with efficient sorting processes, are able to support input materials for high value recycling. However, as mentioned above, the higher the diffusivity of the polymer and, therefore, higher sorption of post-consumer substances the more important is the source control in order to reduce contamination with post-consumer substances or misused packages. The source control can be supplied by modern detecting or sniffing devices which are able to reduce the intake of undesired post-consumer substances into the recycling stream. 23.3.2 Contamination levels and frequency of misuse of recycled plastics Regarding the typical contamination of post-consumer plastics most published data are available for PET bottles and corresponding recyclates. Most of them have quantified or identified substances in post-consumer PET by using different methods. Sadler et al.6,7, published two studies containing data of contaminants in recycled PET. In the first study he pointed out that most compounds found in recycled PET come from PET starting materials, oligomers, flavour bases, label materials and compounds originating in base cups. Contaminants which do not fall into one of these categories are rare. In samples with high levels of contaminants the sum of all compounds was detected to be approximately 25 ppm. No single contaminant appears to be present in post-consumer PET above 1 ppm and all non-usual compounds in post-consumer PET were present below 0.1 ppm. In a second study the identity and origin of contaminants in food grade virgin and commercially washed postconsumer PET flakes were determined. A total of 18 samples of post-consumer recycled PET flakes was examined. In most cases, positive identification was possible, however, in few cases ambiguity resulted from the similarities in mass 500 Novel food packaging techniques
Recycling packaging materials 501 spectra of closely related compounds. Compounds identified were classified into categories associated with their chemical nature or presumed origins, e.g small and ethylene glycol related compounds (methanol, formic acid acetaldehyde, acetic acid), flavour compounds (limonene), benzoic acid or related benzene dicarboxylic acid substances(benzoic and terephthalic acid and corresponding esters, benzaldehyde, phthalates), aliphatic hydrocarbons and acids as well as unexpected and miscellaneous compounds(Tinuvin, nicotine) Bayer has analysed samples from five different recovery systems including PET containers from non-food applications. In these samples he identified 121 substances. The total concentration of all substances found in deposit material was 28.5 ppm. The corresponding concentrations of PET flakes coming from non-food applications were found to be 39 ppm. The key compounds identified were hexanal, benzaldehyde, limonene, methyl salicylate and 5-iso-propyl-2 methylphenol(the flavour compound carvacrol). In conventional washed flakes a maximum concentration of 18 ppm for limonene was determined. For PET flakes from non-food applications the major compound methyl salicy late was determined in a maximum concentration of 15.3 ppm. Additionally the material was analysed after a super-clean process. No peak could be detected in concentrations above the FDA threshold of regulation limit of 0.22 ppm All three published studies mentioned above found no hints for misuse of post-consumer PET bottles e.g. for storage of household cleaners etc. This is most probably due to the fact that these studies are based only on very small amounts of different flake samples. From a statistical point of view flakes from misused bottles should be extremely rare due to high dilution with non-misused PET bottles. Therefore, these published studies are not able to detect the frequency of misuse in typical post-consumer PET flakes In 2002 an EU project under the co-ordination of Fraunhofer IVV was finished.8,%, 10 Within this study 689 post-consumer PET flake samples from commercial washing plants were collected between 1997 and 2001. The samples are conventionally recycled deposit and curbside fractions collected in twelve European countries. In addition, 38 reprocessed pellet samples and 142 samples from super-clean recycling processes were collected. All samples were screened for post-consumer substances, and for hints of possible misuse of the PET bottles by the consumer, in order to get an overview of the quality of commercially recycled post-consumer PET. As a result the average concentrations in 689 PET flake samples for typical post-consumer compounds like limonene and acetaldehyde are 2.9 ppm and 18.6 ppm, respectively. A maximum concentration of approximately 20 ppm of limonene and 86 ppm for acetaldehyde could be determined, which is in close agreement with the above mentioned studies. The impact of the recovery system and the country, where st-consumer pet bottles were collected. on the nature and extent of adventitious contaminants was found not to be significant. However in three bottle flakes hints for a possible misuse of PET bottles e.g. for storage of household chemicals or fuels were found. From a statistical evaluation 0.03 to 0.04% of the pet bottles might be misused. Under consideration of the dilution
spectra of closely related compounds. Compounds identified were classified into categories associated with their chemical nature or presumed origins, e.g. small and ethylene glycol related compounds (methanol, formic acid, acetaldehyde, acetic acid), flavour compounds (limonene), benzoic acid or related benzene dicarboxylic acid substances (benzoic and terephthalic acid and corresponding esters, benzaldehyde, phthalates), aliphatic hydrocarbons and acids as well as unexpected and miscellaneous compounds (Tinuvin, nicotine). Bayer4 has analysed samples from five different recovery systems including PET containers from non-food applications. In these samples he identified 121 substances. The total concentration of all substances found in deposit material was 28.5 ppm. The corresponding concentrations of PET flakes coming from non-food applications were found to be 39 ppm. The key compounds identified were hexanal, benzaldehyde, limonene, methyl salicylate and 5-iso-propyl-2- methylphenol (the flavour compound carvacrol). In conventional washed flakes a maximum concentration of 18 ppm for limonene was determined. For PET flakes from non-food applications the major compound methyl salicylate was determined in a maximum concentration of 15.3 ppm. Additionally the material was analysed after a super-clean process. No peak could be detected in concentrations above the FDA threshold of regulation limit of 0.22 ppm. All three published studies mentioned above found no hints for misuse of post-consumer PET bottles e.g. for storage of household cleaners etc. This is most probably due to the fact that these studies are based only on very small amounts of different flake samples. From a statistical point of view flakes from misused bottles should be extremely rare due to high dilution with non-misused PET bottles. Therefore, these published studies are not able to detect the frequency of misuse in typical post-consumer PET flakes. In 2002 an EU project under the co-ordination of Fraunhofer IVV was finished.8,9,10 Within this study 689 post-consumer PET flake samples from commercial washing plants were collected between 1997 and 2001. The samples are conventionally recycled deposit and curbside fractions collected in twelve European countries. In addition, 38 reprocessed pellet samples and 142 samples from super-clean recycling processes were collected. All samples were screened for post-consumer substances, and for hints of possible misuse of the PET bottles by the consumer, in order to get an overview of the quality of commercially recycled post-consumer PET. As a result the average concentrations in 689 PET flake samples for typical post-consumer compounds like limonene and acetaldehyde are 2.9 ppm and 18.6 ppm, respectively. A maximum concentration of approximately 20 ppm of limonene and 86 ppm for acetaldehyde could be determined, which is in close agreement with the abovementioned studies. The impact of the recovery system and the country, where the post-consumer PET bottles were collected, on the nature and extent of adventitious contaminants was found not to be significant. However in three bottle flakes hints for a possible misuse of PET bottles e.g. for storage of household chemicals or fuels were found. From a statistical evaluation 0.03 to 0.04% of the PET bottles might be misused. Under consideration of the dilution Recycling packaging materials 501
502 Novel food packagi of the PET flakes during washing and grinding with non-misused PET bottles concentrations of 1. 4 to 2.7 misused Pet bottles were estimated from the experimental data. These concentrations can be considered as a basis for the design of challenge tests with respect to sufficiently high input concentrations of surrogates The frequency of misuse was also detected by two other studies. Allen and Blakistone indicate that hydrocarbon for refillable PET bottles ejected between 0.3 and 1% of PET bottles as contaminated. The majority of these rejections came from PET containers with'exotic beverages and not from harmful contaminants. Therefore the part of misused bottles on the rejection in the device is less than 0. 3 to 1%. Bayer et al. -reported the frequency of misuse of pet bottles is one misused bottle out of 10000 uncontaminated bottles. Both studies are in agreement with the results of the EU project In conclusion for PET the predominating polymer unspecific contaminants oft drink ole. Pet contaminants such as phthalates are found far below I ppm. Misuse of PET bottles occurs only in a very low incidence and due to dilution with non- contaminated material the average concentration of substances originated from misuse are also in the lower ppm range. It should be noted here, that the given conclusions are only for PET bottles. If closed loop recycling of other packaging plastics is to be established similar studies on the input concentrations of post consumer substances should be done Comprehensive studies on the contamination of other polymers than PET are vary rare in the literature. Huber and Franzinvestigated 21 reprocessed HDPE pellet samples from the bottle fraction of household waste collections from five different sources. Aim of the study was to investigate the quality of the recycled HDPE samples focusing on substances which are not present in virgin polymers The samples are recycled with conventional washing and extruding steps without a further deep cleansing recycling process. They found that the post consumer related substances in these different samples were the same. They dentified 74 substances which occur in concentrations in the polymers above 0.5 ppm. The predominant species are ester from saturated fatty acids and phthalates, hydrocarbons, preservatives, monoterpenes and sesquiterpenes including their derivatives most of the substances are identified as constituents from personal hygiene products, cosmetics and cleaning agents which are sorbed into the polymer material during storage. The highest concentrations were found for limonene, diethy hexyl phthalate and the isopropyl esters of myristic and palmitic acid, which are present in the concentration range of 50 ppm to 200 ppm. Many odour compounds and preservatives are determined in concentrations of about 0.5 ppm and 10 ppm. They came to the conclusion that due to the concentration and nature of contaminants found in the investigated HDPE samples the recycled material is suitable only for non-food In a second study Huber and Franz investigated a total amount 79 polymer samples(HDPE, PP, PS and PET)from controlled recollecting sources. As a
of the PET flakes during washing and grinding with non-misused PET bottles average concentrations of 1.4 to 2.7 ppm for conspicuous substances from misused PET bottles were estimated from the experimental data. These concentrations can be considered as a basis for the design of challenge tests with respect to sufficiently high input concentrations of surrogates. The frequency of misuse was also detected by two other studies. Allen and Blakistone11 indicate that hydrocarbon ‘sniffers’ for refillable PET bottles rejected between 0.3 and 1% of PET bottles as contaminated. The majority of these rejections came from PET containers with ‘exotic’ beverages and not from harmful contaminants. Therefore the part of misused bottles on the rejection in the ‘sniffer’ device is less than 0.3 to 1%. Bayer et al. 12 reported the frequency of misuse of PET bottles is one misused bottle out of 10 000 uncontaminated bottles. Both studies are in agreement with the results of the EU project. In conclusion for PET the predominating polymer unspecific contaminants are soft drink components where limonene plays a key role. PET unspecific contaminants such as phthalates are found far below 1 ppm. Misuse of PET bottles occurs only in a very low incidence and due to dilution with noncontaminated material the average concentration of substances originated from misuse are also in the lower ppm range. It should be noted here, that the given conclusions are only for PET bottles. If closed loop recycling of other packaging plastics is to be established similar studies on the input concentrations of postconsumer substances should be done. Comprehensive studies on the contamination of other polymers than PET are vary rare in the literature. Huber and Franz13 investigated 21 reprocessed HDPE pellet samples from the bottle fraction of household waste collections from five different sources. Aim of the study was to investigate the quality of the recycled HDPE samples focusing on substances which are not present in virgin polymers. The samples are recycled with conventional washing and extruding steps without a further deep cleansing recycling process. They found that the postconsumer related substances in these different samples were the same. They identified 74 substances which occur in concentrations in the polymers above 0.5 ppm. The predominant species are ester from saturated fatty acids and phthalates, hydrocarbons, preservatives, monoterpenes and sesquiterpenes including their derivatives. Most of the substances are identified as constituents from personal hygiene products, cosmetics and cleaning agents which are sorbed into the polymer material during storage. The highest concentrations were found for limonene, diethylhexyl phthalate and the isopropyl esters of myristic and palmitic acid, which are present in the concentration range of 50 ppm to 200 ppm. Many odour compounds and preservatives are determined in concentrations of about 0.5 ppm and 10 ppm. They came to the conclusion that due to the concentration and nature of contaminants found in the investigated HDPE samples the recycled material is suitable only for non-food packaging. In a second study Huber and Franz14 investigated a total amount 79 polymer samples (HDPE, PP, PS and PET) from controlled recollecting sources. As a 502 Novel food packaging techniques
Recycling packaging materials 503 result they found limonene in nearly all polymer samples independent of the polymer type in concentrations up to 100 ppm for polyolefines(HDPE and PP) and 12 ppm and 3 ppm for Ps and PET, respectively. Limonene can be considered as a marker substance for post-consumer polymers. It is interesting to note that the differences in the limonene concentration are in line with the diffusion behaviour of the polymers. In addition to limonene they found phthalates esters, alkanes, 2, 6-di-tert-butyl-4-hydroxytoluene and oligomers but no hints for misuse of the bottles for storage of toxic chemicals. They concluded that most of the investigated(conventionally recycled) polymers are excluded from closed loop recycling due to the fact that in the polymers substances can be detected which are not in compliance with the European positive list system. It should be noted here that this is an inherent problem of positive lists in view of food law compliance of recycled polymers as well as virgin polymers. A threshold of regulation concept should offer a solution of assuming that a certain concentration of non-regulated compounds is of no concern for consumers 23.3.3 Recvcling technology Today a considerable diversity in recycling technologies can be found, although all of them have the same objective which is to clean up post-consumer plastics Most of them first use a water-based washing step to reduce surface contamination and to wash off dirt. labels and clues from the labels. the material is also ground to flakes during one of the first steps in the recycling process. In most cases these washing steps are combined with separating steps where different materials like polyolefines of PET are separated due to their density. It is obvious, that the cleaning efficiency of these washing processes is normally very different, depending on time, on hot or cold water-based washing or depending on the detergents added to the washing solution. However, typical washing processes are able to remove only contaminants from the surface of the polyme hey are not able to remove organic substances which have migrated in the polymer. Therefore the purity of washed flakes is usually not suitable for closed-loop recycling. A simple remelting or re-extrusion of the washed fakes has an additional cleaning effect, however the purity is usually not sufficient for reuse in the sensitive area of food packaging So-called super-clean processes for closed-loop recycling of packaging materials therefore use further deep cleansing steps. Although there are many technologies commercially available the deep cleansing processes normally use heat and temperature, vacuum or surface treatment with chemicals for a certain time to decrease the concentration of unwanted substances in the polymers. The research on the cleaning efficiency of such super-clean recycling processes has shown that the existing recycling technologies are distinct in terms of rejection of unsuitable material, removal of contaminants and dilution with virgin material. Each of these stages in recycling uses special processes which have an effect on the quality of the finished recyclate containing packaging
result they found limonene in nearly all polymer samples independent of the polymer type in concentrations up to 100 ppm for polyolefines (HDPE and PP) and 12 ppm and 3 ppm for PS and PET, respectively. Limonene can be considered as a marker substance for post-consumer polymers. It is interesting to note that the differences in the limonene concentration are in line with the diffusion behaviour of the polymers. In addition to limonene they found phthalates esters, alkanes, 2,6-di-tert-butyl-4-hydroxytoluene and oligomers but no hints for misuse of the bottles for storage of toxic chemicals. They concluded that most of the investigated (conventionally recycled) polymers are excluded from closed loop recycling due to the fact that in the polymers substances can be detected which are not in compliance with the European positive list system. It should be noted here that this is an inherent problem of positive lists in view of food law compliance of recycled polymers as well as virgin polymers. A threshold of regulation concept should offer a solution of assuming that a certain concentration of non-regulated compounds is of no concern for consumers’ health. 23.3.3 Recycling technology Today a considerable diversity in recycling technologies can be found, although all of them have the same objective which is to clean up post-consumer plastics. Most of them first use a water-based washing step to reduce surface contamination and to wash off dirt, labels and clues from the labels. The material is also ground to flakes during one of the first steps in the recycling process. In most cases these washing steps are combined with separating steps where different materials like polyolefines of PET are separated due to their density. It is obvious, that the cleaning efficiency of these washing processes is normally very different, depending on time, on hot or cold water-based washing or depending on the detergents added to the washing solution. However, typical washing processes are able to remove only contaminants from the surface of the polymers.15,16 They are not able to remove organic substances which have migrated in the polymer. Therefore the purity of washed flakes is usually not suitable for closed-loop recycling. A simple remelting or re-extrusion of the washed fakes has an additional cleaning effect,17 however the purity is usually not sufficient for reuse in the sensitive area of food packaging. So-called super-clean processes for closed-loop recycling of packaging materials therefore use further deep cleansing steps. Although there are many technologies commercially available the deep cleansing processes normally use heat and temperature, vacuum or surface treatment with chemicals for a certain time to decrease the concentration of unwanted substances in the polymers. The research on the cleaning efficiency of such super-clean recycling processes has shown that the existing recycling technologies are distinct in terms of rejection of unsuitable material, removal of contaminants and dilution with virgin material. Each of these stages in recycling uses special processes which have an effect on the quality of the finished recyclate containing packaging. Recycling packaging materials 503
504 Novel food packaging techniques 23.4 Testing the safety and quality of recycled material 23. 4.1 Challenge test The cleaning efficiency of super-clean processes is usually determined by challenge test. This challenge test is based on an artificial contamination of the nput material going into the recycling process. Drawing up a worst-case scenario this challenge test simulates the possible misuse of the containers for the storage of household or garden chemicals in plastic containers. The first recommendations for such a challenge test are coming from the american Food nd Drug Administration(FDA) 8, I9 in 1992. It was a very pragmatic approach The FDa originally suggested realistic contaminants like chloroform, diazinon, gasoline, lindane, and disodium monomethyl arsenate for use in challenge tests However it has been shown in the past that the stability of these surrogates during recycling is in some cases not sufficient. Also the analytical methods in order to detect the surrogates are often difficult to establish and have high detection limits. It is easy to understand that the surrogates used in a challenge est should not degrade during all recycling steps. Otherwise the cleaning efficiency will be better than reality, with adverse consequences towards consumers' safety In the last ten years the selection of the surrogates has moved to chemicals ith more model character. This development was supported by the fact that the range of chemicals available to the customers is extremely limited in practice especially in the case of known genotoxic carcinogens. The surrogates used today in challenge tests cover the whole range of physical properties like polarity and volatility as well as the chemical nature of the compound Additionally, in some surrogates very aggressive chemicals towards the polymer are introduced. However, if too aggressive chemicals are used the physical properties of the polymer and the diffusion behaviour might be changed, which reduces the perception of the challenge test. Nowadays volatile chemicals like toluene. chlorobenzene. chloroform or l1.1-trichloroethane as well as non volatile substances like phenyl cyclohexane, methyl stearate, tetracosane benzophenone, methyl salicylate and methyl stearate are typically used. Of course, other substances with defined physical and chemical properties can be used for a challenge test. It should be kept in mind that such a test should challenge the recycling process in a worst-case scenario. If the resultant ecyclate meets the food law requirements even under such a worst case scenario the process is able to produce recyclates suitable for reuse in packaging applications. In the last decade there have been controversial discussions between scientists, industry and authorities, in view of the worst-case character of such challenge tests. In most cases these discussions arise from the lack of information about the average contamination in the input materials for recycling As mentioned above, the worst-case scenario depends on the concentrations of undesired substances in the post-consumer plastics as well as the frequency of misuse of plastic containers. With knowledge of contamination appropriate safety margins for each polymer type can be defined
23.4 Testing the safety and quality of recycled material 23.4.1 Challenge test The cleaning efficiency of super-clean processes is usually determined by challenge test. This challenge test is based on an artificial contamination of the input material going into the recycling process. Drawing up a worst-case scenario this challenge test simulates the possible misuse of the containers for the storage of household or garden chemicals in plastic containers. The first recommendations for such a challenge test are coming from the American Food and Drug Administration (FDA)18,19 in 1992. It was a very pragmatic approach. The FDA originally suggested realistic contaminants like chloroform, diazinon, gasoline, lindane, and disodium monomethyl arsenate for use in challenge tests. However it has been shown in the past that the stability of these surrogates during recycling is in some cases not sufficient. Also the analytical methods in order to detect the surrogates are often difficult to establish and have high detection limits. It is easy to understand that the surrogates used in a challenge test should not degrade during all recycling steps. Otherwise the cleaning efficiency will be better than reality, with adverse consequences towards consumers’ safety. In the last ten years the selection of the surrogates has moved to chemicals with more model character. This development was supported by the fact that the range of chemicals available to the customers is extremely limited in practice, especially in the case of known genotoxic carcinogens. The surrogates used today in challenge tests cover the whole range of physical properties like polarity and volatility as well as the chemical nature of the compounds. Additionally, in some surrogates very aggressive chemicals towards the polymer are introduced. However, if too aggressive chemicals are used the physical properties of the polymer and the diffusion behaviour might be changed, which reduces the perception of the challenge test. Nowadays volatile chemicals like toluene, chlorobenzene, chloroform or 1,1,1-trichloroethane as well as nonvolatile substances like phenyl cyclohexane, methyl stearate, tetracosane, benzophenone, methyl salicylate and methyl stearate are typically used. Of course, other substances with defined physical and chemical properties can be used for a challenge test. It should be kept in mind that such a test should challenge the recycling process in a worst-case scenario. If the resultant recyclate meets the food law requirements even under such a worst case scenario the process is able to produce recyclates suitable for reuse in packaging applications. In the last decade there have been controversial discussions between scientists, industry and authorities, in view of the worst-case character of such challenge tests. In most cases these discussions arise from the lack of information about the average contamination in the input materials for recycling. As mentioned above, the worst-case scenario depends on the concentrations of undesired substances in the post-consumer plastics as well as the frequency of misuse of plastic containers. With knowledge of contamination appropriate safety margins for each polymer type can be defined. 504 Novel food packaging techniques
Recycling packaging materials 505 23. 4.2 Cleaning efficiency of conventional recycling processes Post-consumer PET which are going into packaging applications are usually recycled with super-clean recycling processes. However, these processes use conventional washing steps as well as several deep-cleansing steps in order to eliminate undesired post-consumer substances from the PET polymer matrix Therefore the cleaning efficiency of conventional washing processes is of interest because it influences the input concentration of post-consumer substances in feedstock material going into the deep-cleansing processes In the literature there are a few studies on the cleaning efficiency of ntional recycling processes. These processes contain washing and surface steps followed in some cases by remelting of the post-consumer I. Komolprasert and Lawson determined the influence of NaOH concentration, mixer speed and temperature on removal of the surrogate tetracosane from spiked PET. In this study percentages of residual tetracosane in the Pet flakes which were washed in small-scale experiments using 13 different conditions were determined. The results show that the tetracosane concentration in the washed flakes was 1. 4 to 3.3% of the initial spiked level As a result only mixer speed and temperature showed a significant effect on removal of the surrogate tetracosane from the PET flakes, while the effect of NaoH concentration was insignificant. The percentage of non-volatile hydrocarbon residues in washed PET flakes varies with the initial concentration. The study determined a removal of 89 to 97% of each hydrocarbon by washing. In a second study Komolprasert and Lawson determined the effect of washing and drying on the removal of surrogates in spiked PET flakes as well as in spiked PET bottles. They concluded that the combination of washing and drying removes 97 to 99% of the organic surrogates from the spiked PET bottles. The copper concentration was found to be 21% of the initial concentration after washing and drying(remark: the low cleaning efficiency for the copper organic compound is most probably due to the instability of this surrogate. It reacts during recycling to Cuo which cannot be removed. This behaviour shows that metal organic compounds are in general unsuitable as surrogates for challenge tests). In case of spiked PET flakes washing and drying removes more than 99% of the initial concentration of the organic surrogates. The high cleaning efficiencies of conventional washing and drying processes are most probably due high temperatures applied during the rying step and due to the fact that contaminants rarely penetrate more than a few u m into the polymer surface. This is in agreement with the result that the initial concentrations of the surrogates in spiked bottles are much lower than those in flakes, because the surface area of flakes is higher than in bottles. A third study from Komolprasert et al. evaluates the decontamination by remelting in a laboratory extruder. The results show that remelting can further reduce the contamination of spiked PET. However, from the data given in this paper, the amount of this reduction is very difficult to evaluate, because of the fact that the some of the applied surrogates(diazinon, malathion, metal organic copper compound)are not stable during extrusion. In addition volatile
23.4.2 Cleaning efficiency of conventional recycling processes Post-consumer PET which are going into packaging applications are usually recycled with super-clean recycling processes. However, these processes use conventional washing steps as well as several deep-cleansing steps in order to eliminate undesired post-consumer substances from the PET polymer matrix. Therefore the cleaning efficiency of conventional washing processes is of interest because it influences the input concentration of post-consumer substances in feedstock material going into the deep-cleansing processes. In the literature there are a few studies on the cleaning efficiency of conventional recycling processes. These processes contain washing and surface drying steps followed in some cases by remelting of the post-consumer material. Komolprasert and Lawson15 determined the influence of NaOH concentration, mixer speed and temperature on removal of the surrogate tetracosane from spiked PET. In this study percentages of residual tetracosane in the PET flakes which were washed in small-scale experiments using 13 different conditions were determined. The results show that the tetracosane concentration in the washed flakes was 1.4 to 3.3% of the initial spiked level. As a result only mixer speed and temperature showed a significant effect on removal of the surrogate tetracosane from the PET flakes, while the effect of NaOH concentration was insignificant. The percentage of non-volatile hydrocarbon residues in washed PET flakes varies with the initial concentration. The study determined a removal of 89 to 97% of each hydrocarbon by washing. In a second study Komolprasert and Lawson16 determined the effect of washing and drying on the removal of surrogates in spiked PET flakes as well as in spiked PET bottles. They concluded that the combination of washing and drying removes 97 to 99% of the organic surrogates from the spiked PET bottles. The copper concentration was found to be 21% of the initial concentration after washing and drying (remark: the low cleaning efficiency for the copper organic compound is most probably due to the instability of this surrogate. It reacts during recycling to CuO which cannot be removed. This behaviour shows that metal organic compounds are in general unsuitable as surrogates for challenge tests). In case of spiked PET flakes washing and drying removes more than 99% of the initial concentration of the organic surrogates. The high cleaning efficiencies of conventional washing and drying processes are most probably due high temperatures applied during the drying step and due to the fact that contaminants rarely penetrate more than a few m into the polymer surface. This is in agreement with the result that the initial concentrations of the surrogates in spiked bottles are much lower than those in flakes, because the surface area of flakes is higher than in bottles. A third study from Komolprasert et al. 17 evaluates the decontamination by remelting in a laboratory extruder. The results show that remelting can further reduce the contamination of spiked PET. However, from the data given in this paper, the amount of this reduction is very difficult to evaluate, because of the fact that the some of the applied surrogates (diazinon, malathion, metal organic copper compound) are not stable during extrusion. In addition volatile Recycling packaging materials 505
606 Novel food packaging techniques substances like toluene are almost completely removed during washing, so that an evaluation of the cleaning effect during remelting on basis of these surrogates is impossible In conclusion, conventional washing processes are able to reduce the input oncentrations of post-consumer substances in flakes. The washing process itself most probably removes only contaminants on the surface of the flakes whereas thermal drying processes are also able to decrease substance which sorbed into the flakes. Remelting processes further reduce the contamination. Due to the fact that conventional recycling processes use a wide range of parameters and quipment, a general conclusion and a quantification of the cleaning effects for washing, drying and remelting processes is not possible on basis of the above mentioned results 23.4.3 Cleaning efficiency of super-clean processes In the last decade studies were undertaken to quantify the residual amounts of chemical substances in the material after deep-cleansing. Therefore the cleaning efficiency of super-cleaning recycling processes is well known Additionally to the challenge test, the quality assurance of post-consumer ecycle(PCR) PET is based on a feedstock control and an analytical quality assurance. Literature data about cleaning efficiencies of super-clean recycling processes are very rare. Three studies of the cleansing efficiency of super-clean recycling processes for PET are published by Franz and process investigated in the first two studies" contains the key steps: washing, re-extrusion and solid state polycondensation(SSP). The process was challenged with three different surrogate concentration levels. as a result the cleaning efficiencies for the different surrogates and contamination levels are between 94 and 99%. The most challenging substance is benzophenone. The results show no significant dependency on the input concentration of the surrogates going into the process. It should be noted here, that this process was tested without a washing process. Including a conventional washing process, the cleaning efficiencies are increased to more than 99.3% even for benzophenone. In the third study" a recycling process without solid stating was investigated. Except for benzophenone, the investigated recycling process reduces all surrogates by more than 95% for initial concentrations below 100 ppm and more than 90% for initial concentrations between 100 and 500 ppm. For that most challenging substance, benzophenone, a cleaning efficiency of approximately 77% at an initial contamination level of 294 ppm was obtained. In conclusion the determined cleaning efficiencies are lower than those of processes with solid stating. However, the specific migration of all surrogates from PET bottles made from contaminated and recycled PET was detected to be far below the migration limit of 10 ppb
substances like toluene are almost completely removed during washing, so that an evaluation of the cleaning effect during remelting on basis of these surrogates is impossible. In conclusion, conventional washing processes are able to reduce the input concentrations of post-consumer substances in flakes. The washing process itself most probably removes only contaminants on the surface of the flakes whereas thermal drying processes are also able to decrease substance which sorbed into the flakes. Remelting processes further reduce the contamination. Due to the fact that conventional recycling processes use a wide range of parameters and equipment, a general conclusion and a quantification of the cleaning effects for washing, drying and remelting processes is not possible on basis of the above mentioned results. 23.4.3 Cleaning efficiency of super-clean processes In the last decade studies were undertaken to quantify the residual amounts of chemical substances in the material after deep-cleansing. Therefore the cleaning efficiency of super-cleaning recycling processes is well known. Additionally to the challenge test, the quality assurance of post-consumer recycle (PCR) PET is based on a feedstock control and an analytical quality assurance. Literature data about cleaning efficiencies of super-clean recycling processes are very rare. Three studies of the cleansing efficiency of super-clean recycling processes for PET are published by Franz and Welle.20,21,22 The process investigated in the first two studies20,21 contains the key steps: washing, re-extrusion and solid state polycondensation (SSP). The process was challenged with three different surrogate concentration levels. As a result the cleaning efficiencies for the different surrogates and contamination levels are between 94 and 99%.20 The most challenging substance is benzophenone. The results show no significant dependency on the input concentration of the surrogates going into the process. It should be noted here, that this process was tested without a washing process. Including a conventional washing process, the cleaning efficiencies are increased to more than 99.3% even for benzophenone. 21 In the third study22 a recycling process without solid stating was investigated. Except for benzophenone, the investigated recycling process reduces all surrogates by more than 95% for initial concentrations below 100 ppm and more than 90% for initial concentrations between 100 and 500 ppm. For that most challenging substance, benzophenone, a cleaning efficiency of approximately 77% at an initial contamination level of 294 ppm was obtained. In conclusion the determined cleaning efficiencies are lower than those of processes with solid stating. However, the specific migration of all surrogates from PET bottles made from contaminated and recycled PET was detected to be far below the migration limit of 10 ppb. 506 Novel food packaging techniques