Polymer Recycling: A Comprehensive Overview and Future Outlook

Abstract

Polymer recycling is an essential and crucial topic in our sustainability-driven society. The depletion of oil and the increasing interest in biomass conversion clearly stimulate the search for alternative carbon sources. On the other hand, polymeric products (plastic, rubber etc.) are ubiquitous and are an integral part of our life. Recycling these products is thus of paramount importance, and perhaps crucially, from an environmental point of view. In this work, we will focus on the most common commodities, with the most important being (in terms of production volume) thermoplastics, rubbers and thermosets. A consequence of this choice is that the most common materials as well as chemical and biochemical recycling methods will be discussed. New advances in the corresponding technologies will be presented and critically evaluated. Finally, on the basis of this literature review, we will identify current trends and possible future developments.

1. General Definitions and Aim of This Work

The recycling of polymeric products is a necessity in the current sustainability-driven climate of our society. This has stimulated an increasing number of research projects, which at the academic level has resulted in an exponential increase in publications during the last three decades (Figure 1).

This general topic and its popularity, as the objective of scientific and technological development, stem among other environmental concerns as well as economic considerations conceptually linked to the depletion of fossil resources and the fate of polymeric products at the end of their lifetime [1].

The legislation in European countries clearly recommends the re-use and recycling of polymer waste (as opposed to landfill) as its main priorities [2]. On a longer timescale, biodegradation would obviously be the preferred route [3], even though the task of replacing current plastics with fully biodegradable ones can only be defined as aspirational. When dealing with recycling, which is often seen as the main solution to decrease the environmental impact of polymeric products [4,5] as well as decrease the energy demand [6], multiple options are available (Figure 2) and constitute the objective of many multidisciplinary research projects [7] where technical aspects are often combined with economical and environmental ones [8,9].

We notice here that both the materials and chemical recycling offer side opportunities as ground waste (see below), while chemical recycling often results in oil that can be conveniently used as fuels and fillers for composites.

We use here the term material recycling as opposed to the widespread “mechanical recycling” for two reasons. In the first instance, the word “mechanical” does not encompass the possibility, also when dealing with a single recycled polymer, of chemical reactions during processing (e.g., extrusion). The paradigmatic cases of PP (giving degradation upon recycling) and polycondensates (like nylon, which usually branches during processing) constitute relevant examples [10]. Moreover, as discussed below, in cases where polymer blends are considered, these almost always imply the presence of chemical reactions at the interface with the formation of compatibilizers to improve the dispersion of one component in the other. In this case, among other possibilities, the paradigmatic example of transesterification reactions between PET and PC can be put forward [11]. This also holds true for more complex systems, for example, for textile waste. The blends of polyamide-6, polyether-urea copolymers and rPET showed the relevant influence of interchange reactions between ester-amide groups and possibly additional ones (acidolysis, alcoholysis and aminolysis) during processing in the melt [12].

Thermosets and rubbers deserve special mention as their crosslinked nature factually hinder, although not entirely [13,14,15,16], the very few attempts at material recycling. For rubbers in particular, the use of thermoplastic elastomers constitute a valid alternative, whenever possible, based on the product requirements [17]. For thermosets, at least PU or epoxy resins, the aminolysis of the urethane and C-O bonds, respectively, seems to constitute a viable option [18].

Finally, the crucial difference between industrial and post-consumer waste must be stressed here. The former is relatively clean and pure while the latter suffers from severe contamination from impurities and other plastics, thus being much more difficult to recycle [19].

Generally speaking, material recycling results in worse properties (thermal and mechanical) with respect to the virgin materials [20,21,22,23,24,25,26], even though in specific cases, only at the level of physical appearance [27]. It is stressed that material recycling does not solve the problem of the negative impact of plastics on the environment; this solution only postpones the same consequence. This is because after the end-of-life of the recycled plastics, they are converted again into waste or litter [28]. Indeed, feedstock recycling (either chemical or biochemical) displays the advantage of closing the loop and constitutes a conditio sine qua non as upon multiple recycling steps will ultimately deteriorate the polymer properties in a decisive way [29]. The dichotomy between material/mechanical vs. chemical recycling has been the subject of many controversial contributions. It is sufficient here to say that several life-cycle analyses have shown the convenience of material recycling for commodities such as low-cost polymers (e.g., polyolefins) [30] and thermoplastic composites [31]. This is not surprising when making allowances for the fact that the separation of waste plastics and their processing (via reactive extrusion) are certainly and self-speaking less energy intensive than pyrolysis, which implies of the separation of the monomers from the resulting oil, re-polymerization, followed by an extrusion step to pelletize the new polymer [32], with the last two steps also being carried out when using monomers from fossil resources. As a result, the discussion seems fairly obsolete; although, it should be noted that a definitive advantage of the chemical recycling route is the presence of suitable infrastructure as oil refineries can be conveniently used to refine the pyrolysis oil [33].

In this work, attention will be paid to commercially available polymers, following similar overviews already published in the past [34] in order to provide the reader with a comprehensive summary of the state of the art. In Future Outlook, we provide a concise overview of recent and present developments. Our section on design for recycling will try to forecast possible application of reversible chemistry for the circular economy while the reader is referred to recent works for the more industrial-oriented product design for recycling [35].

2. Recycling of Current Commercially Available Polymers

2.1. Pre-Treatment

The first step in material recycling for post-consumer waste is the sorting of materials into different components. This is conveniently achieved on the basis of density differences but also with spectroscopic (especially FTIR and Raman [36,37]) methods. For example, FTIR has proven useful in separating acrylonitrile-butadiene-styrene (ABS), high-impact polystyrene (HIPS) and ABS/polycarbonate blend (ABS/PC). Fire tests can be conveniently carried out to detect the presence of flame retardants as in ABS and HIPS [36]. This is crucial as the presence of flame retardants can constitute a problem in terms of side-reactions during the melt processing [38]. Generally speaking, sorting still faces many challenges due to the heterogenous nature of the waste and the advent of novel complex products, for example, from the biomedical sector [39]. Separation from metals is also becoming a relevant issue when dealing with recycling of different materials from electric vehicles [40]. Moreover, specific problems might also arise when sorting specific waste. For example, textiles usually consist of different polymeric materials such as polyesters, polyamides and acrylics. Their presence in mixed yarns [41] constitutes a relevant challenge, besides the expected issue of contamination with other compounds [42]. As well as spectroscopic characterization, more specific techniques are often required for a precise determination of the polymer’s physical properties. For example, a novel GPC technique has been reported for characterizing the molecular weight of recycled polymers [43]. This development requires less differences in refractive index values between polymer and solvent/contaminants, which is obviously a common case for waste materials.

After the sorting step and on the basis of different characterization techniques, it might be clear that the quality of the waste is not enough to justify a reprocessing step in the melt. In these cases, grinding into solid particles might still offer an option for re-use. Indeed, the use of bitumen, asphalt and concrete as low-cost application for waste polymers [44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71], especially in the case of crosslinked materials [72], is still quite popular. For example, the use of recycled latexes from the paint industry in concrete delivered the expected effect on the final product, namely improved bond stress in the elastic region [73]. Their use in concrete remains attractive for recovered fibers [56,74] or even single-polymer fiber-reinforced thermoplastic composites [75]. This strategy is quite general and implies the use of recycled polymer in the overall construction industry [76]. It must be stressed here that the complex nature of these materials makes it very difficult, if possible at all, to pinpoint the exact function of the polymers as related to their (chemical) structure. As a consequence, the product designers do not have at their disposal any exact tools for selecting the most suitable materials for a given application, and comprehensive/comparative studies will probably still be needed.

Along the same line, a novel application area is the use of waste polymers in metallurgical and mineral manufacturing processes. This is quite interesting as recycled polymers can be used (for example as interfacial agents and binders) as well as their pyrolysis products (for example H₂, CH₄ and CO), with the latter as reducing agents [77]. The application of recycled polymers as binders has also been described [78].

When potential exists for the direct collection or easy sorting of a plastic waste [79], many technologies become available for recycling [80]. One such example is constituted by ABS from electronic goods. This can be collected and segregated in a relatively easy way according to the existence of proper legislation in many countries. The purified waste obtained can then be recycled mechanically or thermally [81]. For example, it has been shown that ABS from monitors can be efficiently recycled as material with little variation in thermal and mechanical properties [82].

After sorting, a further, often more accurate strategy, for the separation of different polymeric materials, is the use of selective solvents as in the case of PC, PS and ABS [83]. These processes are often quite accurate and able to distinguish quite a number of different polymers, for example, for packaging materials based on LDPE and HDPE, PVC, PS and PET. Generally speaking, high yields (>90%) and purities are achieved [84]. This is popular as it allows in principle multi-layer recycling even if its applicability is dependent on the material under investigation [85]. Furthermore, its environmental impact is questionable as many potentially dangerous and non-environmentally friendly organic solvents are used [86,87]. The same holds true for polymeric products displaying the presence of additives in relevant percentages with PVC being a paradigmatic example [88]. Despite the difficulties in handling organic solvents at large scale, the selective dissolution strategy remains interesting, from an economic perspective [89], for multi-layer films when the objective is the recovery of the single components [90].

2.2. Material Recycling

Material recycling is a very convenient option for industrial waste or, generally, when the purity of the waste material justifies a direct re-use. This hold true also for biodegradable polymers (usually polyesters); material recycling constitutes an added value, besides biodegradability, in terms of end-of-life policy [91]. This has been demonstrated for PLA/PHBV blends [92] and for PLA alone [93] upon the addition of a chain extender [94].

The situation is completely different for post-consumer waste. Indeed, after sorting, the co-existence of physical mixture of polymers is more so the rule rather than the exception. If individual components can be separated, generally speaking, the mixing of a recycled polymer with the virgin one provides ample opportunities for properties control as shown in PE [95,96], ABS [97] and PLA derivatives [98]. This is necessary when the separated and recycled component is not pure enough to be re-used on its own, depending on the envisioned application. In this context, we note that there is not a general rule in order to assess whether the recycled polymer is feasible for a general application or not. It is necessary in every specific case to conceptually retrieve the specific product requirements and, for example, use product design tools [99], such as the house of quality, to check for the possibility of re-using the polymer without any extra processing step. Multiple separation and purification steps might compromise the economic feasibility and possibly imply the use of hazardous chemicals and not sustainable technologies.

If further separation is not possible or feasible from an economic point of view, a logical solution would actually be to blend the polymers in the melt. Generally speaking, the potential of polymer blends (as originated from unsorted plastic waste) is quite underestimated [100,101]. This ultimately stems from, among others, the over-design nature of many polymeric products. Indeed, the use of a specific material for a given application has been driven, during the last few decades, by technical (satisfaction of the product requirements) as well as economic considerations (economic feasibility) [102,103,104,105,106,107,108,109]. The availability of many materials that satisfies product requirements and are commercially available at relatively low price also led to over-design. Examples of the blending strategy can be found in PLA/rHDPE [110], rPET/PC [11], rSAN/PVC and rABS/PVC [111] blends and even for rubbers such as EPDM/rEPDM [112]. This strategy often results in products with (slight) inferior quality, thus suggesting a down-cycling characteristic [113]. Obviously, one tries to compensate for the loss of properties in the recycled material with the gain obtained by choosing a suitable second component in the blend. An example is constituted by rPET/PC blends where the relatively high T_g and barrier properties of the PC are crucial in compensating for qualities of the rPET [11]. Special cases can be highlighted when the waste originates from a given source and with a known composition. For example, polymer blends based on 20 different waste streams of printers have been reported to consist of HIPS/PS (90/10 wt/wt), HIPS/ABS (90/10 wt/wt) and pure HIPS. Despite having slightly different mechanical properties as a result of the different chemical structures, materials obtained by formulating the different components displayed consistent mechanical behavior and could be used to manufacture hangers, organizing boxes, soles and watering cans [114].

The blending of different components (usually present already as physical mixtures) represent a valid option especially when suitable compatibilization techniques are used (for example PET/PE/PC, Polyolefins/Nylon or PET blends [115] and even quaternary ones consisting of PE, PP, PS and PET [116]). The compatibilization of polymer blends is most often a pre-requisite to obtain the desired mechanical behavior. This necessity stems from the immiscible nature of polymers [117,118] and thus the necessity to improve the dispersion of one component into the other. Compatibilization helps in strengthening the interfacial adhesion between different polymeric components. This also helps when dealing with recycled fibers that most often lack interfacial interaction with polymeric materials, thus resulting in weaker composites [119]. This might be balanced using surface treatment. One example is provided by using plasma treatment to modify the surface properties for PET films used as adhesive [120] but also for carbon fibers [121]. Finally, the addition of specific reactive monomers (e.g., acrylics) has been reported to also increase the interfacial adhesion between PP and different kind of fibers [122].

Compatibilizers are traditionally prepared in the melt via the functionalization of virgin polymers (Figure 3) [123].

We take here the functionalization of iPP with MAH as a paradigmatic example even if the chemistry of these processes is heavily dependent on the substrate and the used monomers [123]. These reactions and the function from the corresponding compatibilizer (precursor) date back to the last two decades of the 1900s (see for example [123,124,125,126]) but have been recently “brought back to life” in many different studies (vide supra) [127].

Also, in the case of polyolefin blends, the use of reactive additives, in this peculiar case, peroxide helps in improving the compatibility between different components, such as PE and PP [128] (Figure 4).

This is probably due to the formation of graft/block copolymers (via radical coupling) and the subsequent action of the latter as compatibilizer for the blend. The use of compatibilizers has been reported as crucial in improving the compatibility with solid particles for polyolefins waste from agricultural silage films [129]. The latter is mainly constituted by PE with small PP impurities. However, they also comprise solid particles (sand and minerals) that are not easily recovered after grinding the used films. When recycling the obtained particles via extrusion, the addition of a compatibilizer (typically a maleic anhydride, MAH, functionalized polymer at 2.5 wt % intake), a chemical reaction between the -OH groups at the surface of the solid particles and the MAH, is inferred to take place (Figure 5).

The result of such reactivity is the formation of chemical bonds between the filler and the matrix and consequently the retaining of the mechanical properties compared to the virgin film. This holds true quite generally for all fillers displaying the presence of a surface of suitable (e.g., -OH) reactive groups, thus including polysaccharides (e.g., starch and cellulose), wood, glass, etc. [130,131,132].

The use of compatibilizers is crucial when dealing with polymeric products that comprise a polymer blend and/or a block copolymer [133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149]. This stems from the fact that polymers are not miscible, generally speaking, at molecular level, with this being true also for the blocks of block- and graft-copolymers. The net result of this is the formation of multiphase morphology where one or more components are dispersed into another one functioning as the matrix. When degradation reactions take place during processing (particularly common for polyesters and PP), the formation of low molecular weight chains and shorter blocks for copolymers will generally result in coarser dispersions with relevant changes in the mechanical and rheological properties. The use of a “fresh” compatibilizer is then meant to counteract this effects and provide morphology stabilization upon processing.

The concept of surface reaction can be exploited also for the recycling of crosslinked particles (particles that do not melt upon heating) by creating an interpenetrating network as demonstrated in the case of SAP [150] with silane-coupling agents (Figure 6).

The obtained product has demonstrated a retention of the absorption properties higher than 80%, thus demonstrating the validity of this approach. In reality this type of strategy is widely applied as a kind of last resort (vide supra). When the quality of the waste mixed plastic is not enough to justify reprocessing in the melt, the use of the ground solid as filler in combination with a resin still allows for the production of composites [151,152].

The promotion of radical reactions during extrusion has benefits not only for polymer blends [153,154] but also for homopolymers. As shown for the recycling of polyamides, the use of γ-irradiation might result in the branching/crosslinking of the polymer [155], thus resulting in a viscosity increase, with respect to the virgin polymer. The addition of stabilizers (typically radical quenchers) also represents an effective strategy to counterbalance possible degradation reaction upon recycling [156], even though these low molecular weight compounds do not usually help in improving the compatibility between polymers. In some specific cases, the use of recycled polymers also results in superior properties with respect to the virgin one as in the case of rPP in wood composites. This was attributed to the increased crystallinity of rPP as compared to the original one (probably as a consequence of the thermal degradation). The formation of a transcrystalline layer at the surface of the wood fibers was deemed responsible for the observed mechanical behavior in terms of creep compliance [157].

Food packaging (e.g., PET [158] and polyolefins [159]) requires special regulation to be re-applied in the same field [160]. This often results in the need for dedicated policies for the return of specific waste so as to minimize possible contamination [161]. However, the combination of recycled material with virgin material [162,163] in a multilayer structure could also represent a solution, provided that the migration kinetics of pollutants between the recycled and the virgin layer is known [164,165,166] and possibly controlled [167].

Blending is not the only option for material recycling. From a technological point of view, melt blowing represents a convenient method to directly convert recycled polymers into a fabric/film as already demonstrated for PP [168]. A recent development also include the concept of microfibrils, for example, blends of PET and LDPE where the former is present as microfibrils [169], ultimately displaying properties superior upon cold drawing [28]. Micronization is also an alternative to blending when weaker bonds (for example O-O or S-S) are present [170].

The idea of recovering high-value materials from polymeric products finds a relevant example in the recovery of glass and carbon fibers [171,172,173,174]. In particular, carbon fibers, given their increase in production and broad range of applications, have been receiving significant attention in the past decade [175,176,177,178]. It is worth noticing here that the recovery of these fibers might display quite general benefits [70,119] as their use in composites with recycled polymers as matrix significantly dampens the effect of recycling. This is a simple consequence of the fact that the property of a composite depends also on the property of the filler, according to equations like the one below (1) for the elastic modulus (E).

$$E=E_m\times \left(1-{\varnothing }_f\right)+E_f\times {\varnothing }_f$$

(1)

where E is the elastic modulus; ϕ is the volume fraction, and the subscripts m and f refer to the matrix and filler, respectively. If we take, for example, the elastic modulus of PP (2 GPa) and carbon fibers (170 GPa) [179] at ϕ_f = 0.05, the modulus of the composite is equal to 10.4 GPa. A 30% loss of modules for PP upon recycling would then result in a composite with modulus equal to 9.83 GPa, resulting in a loss of about 6%. Note that Equation (1) (known as the rule of mixtures) only gives the maximum elastic modulus of a fiber composite, even though the composite may show anisotropic behavior due to fiber orientation. In addition, it does not consider the voids or size of the fibers, although these aspects can be considered in other models as well [180]. Natural fibers (such as palm [181,182], silk and cotton [183] ones) can also be used. Recently, the use of nano-additives has been proposed to minimize the property loss [184,185,186,187,188]. Besides fibers, the addition of different fillers (such as zeolites [189], ash [190], SiC/Al₂O₃ particles [191], metal powder [192], recycled materials [193,194,195] and toughening particles [196]) to unsorted polymeric waste has also been reported. Comparative studies are able to highlight the difference between different fillers and in some special cases pinpoint synergy in terms of the mechanical behavior of the composite [197]. In special cases, the presence of the reinforcement also results in improved stability against weathering and aging [198]. All strategies briefly outlined above constitute a part of the sustainability practices that are currently being defined and implemented for composite materials [199,200]. It goes without saying that the possibility of recycling both the matrix and the filler constitutes, whenever possible, the ideal scenario [201].

The general strategy of recovering high-value fibers [202,203], even with non-conventional methods [204], is also valid when using recycled fibers in in situ polymerization processes [205] as additives in concrete [206]. A recycled polymer can also be used to produce fibers [207] and can be then used as reinforcements in new composites as shown for polyamide 6 [208], PP [209] and PET [210].

In case separation is not an option, grinding and use as a filler for the all composite is still an investigated option [211], in addition to mixing with concrete [212,213]. Recently, the possibility for thermoplastic composites of grinding and then application for 3D printing has been demonstrated for wind blades [214] and other polymer waste [215,216] based on PLA [217,218,219,220,221], PET and derivatives [218,219,220,222], such as ABS [218,219,221] HDPE [219]. It should be noted that this approach often relies on a relatively high purity of waste streams, but, whenever possible, it also reduces the energy demand as well as the carbon footprint of the recycling process [223]. The use of recycled fibers in new composites for addictive manufacturing is also gaining a lot of interest as recycled fibers are more cost effective, albeit possessing inferior properties, than virgin ones [224]. The application of recycled composite have also been found in water treatment [225]. Generally speaking, additive manufacturing seems to represent a large application field for many polymers to the point that a specific codification procedure is proposed to keep track of the material used [226].

2.3. Chemical Recycling

The general idea of chemical recycling consists of destroying the polymeric structure, whenever possible via depolymerization reactions, so as to recover the (original) monomers. Pyrolysis, in this context, represents the most used technique especially for polyolefins due to the difficulties in other methods for separation [227,228], especially when blended with other polymers [229].

Chemical recycling offers a promising solution to enhance recycling rates, with microwave heating emerging as an attractive technology for polymer breakdown [230,231,232,233], in order to improve the degradation kinetics. However, the outcome, in terms of chemicals obtained and possibility for direct re-use in polymerization, remains heavily dependent on the complex chemistry of degradation [234] as well as the applicability (also from an economic standpoint) of suitable separation techniques. Indeed, many works refer to the pyrolysis of oil for other applications such as fuels [235].

Also in this case, when the waste has a consistent chemical structure, for example, when originating from a common application, interesting results have been reported. The pyrolysis of crosslinked PMMA (from dental waste) yielded >90 wt % at the liquid phase with >98% purity in the monomer. After purification via distillation and re-polymerisation, the properties are similar to the ones of the virgin polymer [236]. Similar results have also been reported by using innovative technologies such as an indirectly heated fluidized beds [237].

For polycondensates, chemical recycling is undoubtedly easier than for polyolefins or other polymer involving radical chemistry. Indeed, in this special case, the polymer and the corresponding monomers are usually in thermodynamic equilibrium with each other. This allows for the use of reactive chemicals to depolymerize the material to low molecular weight compounds (not necessarily the monomers) that can be re-polymerized again in a theoretically straightforward manner. The approach is quite general (Figure 7) as it applies to a wide range of polycondensates [238,239].

Generally speaking, waste based on thermosets requires destructive approaches so as to destroy the three-dimensional structure [240,241] or, in any case, eliminate the bonds between the chains, similar to what happens in rubber devulcanization [13,14,15]. This also holds true when trying to recover fillers. Indeed, in case of thermosets, special methods are needed to recover the fibers [242], for example, the degradation of matrix using superheated steam [243], supercritical water [244], alcohols [245] or acetic acid near the critical point [246]. Comparative studies have also shown a clear dependence of the composite properties on the strategy to recover the fibers [247]. Interesting developments involve the use of multiple steps to break the thermoset structure and recover the fibers. An example of this is the chemical disruption of an epoxy-based thermoset [248] (Figure 8).

It is believed that the used solvent diffuses inside the composite, thereby causing swelling and increasing the accessibility of the functional groups. The Lewis acid catalyst can then selectively break the C-N bonds, thus effectively degrading the three-dimensional structure to the level of soluble chemicals. This ultimately leaves the fibers behind.

The chemical routes highlighted above become crucial when dealing with crosslinked polymers. Indeed, rubbers and thermosets deserve a special attention as their crosslinked structure does not allow for direct materials recycling. The reader is kindly referred to the many reviews and seminal works on this topic for specific details [249,250,251,252,253,254,255,256,257,258,259,260,261,262,263,264,265,266,267,268,269,270,271,272,273,274,275,276,277,278,279]. We note here that material recycling (e.g., devulcanization for rubbers) has reached quite low TRL levels [249] as it is factually hindered by several factors among which is the costly and often dangerous nature of the devulcanization agents used and the fact that the net result is still a mixture of different rubber polymers with fillers and additives. On the other hand, material recycling in the form of ground solid remains an attractive option for rubbers in concrete [280,281,282,283,284,285,286,287,288], asphalt/bitumen [289,290,291,292,293,294,295] and composites [296,297,298,299,300,301] as well as for thermosets as solid fillers in composites [302]. For both kind of materials (rubbers and thermosets), chemical recycling routes, at the moment, seem to be the preferred strategy [303]. For rubbers, this attractive characteristic is linked to the main product of chemical recycling: on one side, oil that can be conveniently used as fuel [304,305] and a solid residue with practical applications as carbon black [306,307,308].

A more fundamental approach to material recycling for rubbers and thermosets is stringently needed and discussed, among others, in the following section.

3. Future Outlook

The stringent necessity for a circular policy, for example, European countries committing to full circularity before 2050, will most probably and hopefully attract even more interest in polymer recycling in the coming years. In the long term, the design of novel materials, at the molecular level, will allow for easier recycling, as recycling deserves a relevant spotlight. On the other hand, novel developments for currently available materials are also needed. We highlight both strategies in the following section.

3.1. Design for Recycling

New developments are needed to start producing materials which are inherently recyclable at their end-of-life [309]. In the last couple of decades, the development of covalent adaptable networks (CANs), also referred to as dynamic covalent networks, rapidly gained momentum. This is due to their ability to potentially replace the conventional thermosets of which the crosslinks cannot be broken easily. Additionally, these types of networks can sometimes exhibit a remarkable feature of self-healing, which expands the service lifetime of the products. The development of CANs spans from the discovery of new reversible ‘click’ chemistries to the use of these chemistries to design new CANs, and finally tailoring their properties to the product applications. CANs possess reversible covalent bonds which can be broken under a variety of stimuli, among which thermal energy and light are the most common triggers [310]. The equilibrium reactions at the foundation of CANs can be classified into two types, depending on the mechanism of the bond formation and dissociation. As illustrated in Figure 9, bond exchange can occur via an associative or dissociative mechanism. Associative networks (also known as vitrimers) have the ability to undergo bond exchange and thus possess a degree of flow, while still maintaining the network integrity and the number of crosslinks throughout the recycling process. In contrast, dissociative networks lose complete network integrity after thermal treatment. This becomes apparent as thermal treatment results in a lower viscosity as compared to vitrimers. For this reason, dissociative networks are considered to be less challenging to recycle. As illustrated in Figure 9, the reversible bonds can only be part of a crosslink between polymer strands, or the monomers themselves are designed with functional groups to eventually produce a crosslinked network with reversible bonds in the ‘main chain’ as well. We touch upon several considerations in the design of CANs, being the types of chemistry that induces bond dissociation or exchange, several strategies to synthesize such networks and the typical recycling conditions of CANs.

Due to the double step mechanism of dissociative reactions, the crosslink density depends on the time and temperature. Therefore, it will vary during the reprocessing step and commonly thereafter as well [311]. The most common chemistry being employed for the design of dissociative CANs is the Diels–Alder reaction, as displayed in Figure 9B. The furan–maleimide couple in particular has been introduced in polymer networks in several ways, notably by exploiting functionalities in the backbone and introducing furan groups. For instance, poly-ketones can be functionalized with furan groups with a Paal–Knorr reaction with furfuryl amine [312,313]. Others have demonstrated the functionalization of jatropha oil [314,315] with furan groups via epoxidation of unsaturated bonds, followed by an epoxy–amine reaction with furfuryl amine. Similarly, this has been demonstrated for the design of recyclable rubbers [316,317]. Dissociative networks are typically thermally recycled due to macroscopic flow induced by the retro Diels–Alder reaction at elevated temperatures. Higher temperatures break the Diels–Alder crosslinks, and thus, the recycling will be similar to that of a thermoplastic. Typically, the equilibrium conversion of the Diels–Alder reaction starts to decrease at above 70 °C [318]. The actual application window may therefore be narrower compared to the associative networks. The temperature at which the network undergoes a phase transition from a solid to a viscous liquid depends on the crosslink density. This in turn depends on the network architecture and the number of functional groups per monomer. Paul Flory and Walter Stockmayer developed a theory on molecular size distribution and the percolation of multifunctional end-capped branch units, leading to a simple relationship [319,320].

$$x_{gel}={\displaystyle \frac{1}{\sqrt{(1-f_A)\left(1-f_B\right)}}}$$

(2)

Equation (2) is known as the Flory–Stockmayer equation. Here, $x_{gel}$ is the critical gel conversion, i.e., the minimal conversion at which every branch unit becomes part of the total network. This can be calculated with the number of functional groups per monomeric branch unit. $f_A$ and $f_B$ are taken here as the number of functional groups of a furan-bearing monomer and a maleimide one. The equilibrium conversion decreases by heating the material to below $x_{gel}$ to induce macroscopic flow. Flory–Stockmayer theory is in line with experimental ‘de-gelation’ temperatures and can accurately predict at which temperature the gel–liquid transition takes place and thus at which temperature it can be recycled. As such, several publications highlight the good agreement for Diels–Alder systems built from furan end-capped monomers and a bismaleimide [321,322,323,324], which was made possible due to their accurate description of the reaction kinetics. In a dissociative network, the main challenge is to recycle without any side reactions. For Diels–Alder networks, the reaction rate is high enough to push the equilibrium downwards in minutes or even seconds at 140 °C, until the point where perhaps the system is heat-transfer limited instead of kinetically limited. However, at these temperatures some other side reactions can jeopardize the reversibility of the system. The most considered ones are maleimide homo-polymerization [325,326], double Diels–Alder [327] or aromatization [328,329,330], although the given methods to prove aromatization are sometimes debatable. Other Diels–Alder couples involving anthracene have been used as well and are generally more stable at higher temperatures, and for this reason, they also require a higher temperature or time to be reprocessed compared to furan–maleimide systems [331,332]. Other types of chemistries for this purpose also are explored and summarized in Figure 10. For example, Schiff base reactions involve imine bonds that can undergo both an associative pathway (trans-amination) and a dissociative pathway, being a hydrolysis reaction towards the ketone and primary amine [311]. This mostly occurs in acidic conditions and in water, and these materials could be reprocessed via acid-catalyzed degradation in solution to retrieve the starting materials or alternatively by thermal recycling [333,334].

Boronic acids and diols can form a reversible boronic ester bond and emerged as an attractive and safe method for biomedical applications [335]. The complexation of a diol and a boronic acids is pH responsive, and therefore, its reversibility has been mostly demonstrated with pH as a stimulus. Similarly, acylhydrazones are dissociative as well and are recyclable upon changes in pH [336]. pH activation would therefore be favored to induce reversibility as thermal activation brings up concerns about the thermal stability. The conditions used to demonstrate the recyclability of these crosslinked networks are reported in

Table 1. Reported recycling conditions sorted based on the type of reversible dissociative chemistry for comparative purposes.

Chemistry	T (°C)	Method	Comments	Reference
Diels–Alder: furan–maleimide crosslinked castor oil	130	Free flow into mold	three cycles	[337]
Diels–Alder: furan–maleimide crosslinked polyketones	120–150	Dynamic mechanical thermal analysis	seven cycles	[338]
Diels–Alder: furan–maleimide crosslinked EPDM rubber	175	Hot pressing	One cycle shown in tensile tests	[317]
Schiff base: vanillin based	50	Acid hydrolysis	Shown once with NMR	[339]
Schiff base: vanillin based	170	Hot pressing	three cycles	[333]
Oximes	155	Hot pressing	four cycles	[340]

along with the reported recycling conditions.

The mechanism in associative networks requires an additional third party to take place in the exchange process. Heat induces both thermal motions to bring these groups together and faster exchange kinetics [341]. Some exchange reactions, such as trans-esterification, are rather slow and may require a catalyst to stimulate the rapid bond exchange. In their review, Kumar et al. provided an overview of reprocessing conditions for trans-esterification reactions [342]. Transesterification is often catalyzed by zinc acetate, as was conducted in the pioneering work of Montarnal et al. [343], and somewhat lowers the temperature to reprocess the material. Zhang et al. [310] provided an overview of several strategies to recycle vitrimers and provided examples in which conditions where even vitrimers can be continuously recycled. Taplan and co-workers [344] highlighted that, so far, the reprocessability of vitrimers is limited to compression molding only, but they managed to process vitrimers in a continuous fashion as well through extrusion at 150 °C. This was performed by increasing the bond exchange rate through careful network design. Several examples of typically reported recycling conditions are listed in

Table 2. Reported reprocessing conditions for vitrimers, listed for various types of associative chemistries.

Chemistry	Temperature	Method	Comments	Reference
Trans-esterification: fractionated lignin and sebacic acid	160 °C	Hot pressing	Zn(acac)2 as catalyst	[345]
Trans-esterification: palm oil-based epoxy and citric acid	170 °C	Hot pressing	Catalyst free	[346]
Di-sulfide metathesis	180 °C	Hot pressing	Three cycles, welding performance tested	[347]
Polyurethane	150 °C	Extrusion	Twin-screw extrusion, one cycle	[344]

for a variety of reversible chemistries. It should be noted that the conditions to recycle are chosen for the sole purpose to prove the concept of recycling, even though it may be possible at lower temperatures as well. Overall, the thermal cycles, via tensile tests, employed in these examples highlight the recovery of mechanical properties after several recycling steps.

3.2. General Considerations

For currently available materials, many trends and necessities can be identified. In the first instance, further improvements of the sorting process is desirable (for example, based on macroscopic properties such as color and density [348]) in order to be able to more efficiently separate almost pure components. In this respect, advances have been made in improving the spectroscopic techniques, such as FTIR [349]. This is also important at the polymer level as biodegradable polymers (e.g., PLA) are gradually replacing oil-based non-biodegradable ones (e.g., PE) in several applications [350,351]. Being able to identify the presence of both kind of polymers is paramount in defining suitable separation strategies or blending ones in case the mixed waste cannot be separated [352].

Secondly, the design of a high-value application and, in general, upcycling strategies [353,354], for example, in batteries [355], are needed in order to boost research in the field and compensate for the low-value applications. One other way is to modify the recycled polymer (directly during extrusion), for example, by grafting PE with GMA [356,357] (Figure 11).

The modified polymer is obviously more polar than the original PE and as such can be used, for example, for adhesive applications.

Thirdly, improved and more sustainable pre-treatment techniques should be systematically developed. Despite the many advantages of material recycling, as outlined above, one key aspect to be considered is the presence of additives such as stabilizers and plasticizers. If these low molecular weight compounds (for example, brominated flame retardants [358,359]) present threats to peoples’ health or the environment, they have to be removed prior to recycling. The same holds true in cases where these molecules simply interfere with the chemistry, if any, involved in the recycling step in the melt [360]. A new trend is detectable to tackle this problem, namely the use of green solvents as a pre-treatment step. At the same time, comparative studies have been carried out to selectively isolate and separate low molecular weight additives (as well as oligomeric fractions generated via degradation upon recycling) from the bulk-recycled polymers [361].This knowledge is crucial in selecting the most suitable separation process and thus renders the recycled product as pure as possible.

Another aspect deserving attention is the presence of an intermediate method between material and chemical recycling. When degradation cannot be avoided (for example upon multiple recycling steps), the oligomeric nature of the obtained product makes it possible to find applications as lubricants [362]. More importantly, the oligomer’s route seems to be developing due to lower energy demand and the potential to easily return to unprocessed materials [363].

Developments are also needed at a theoretical level, for example, with the use of advanced machine learning techniques, to predict polymer properties when mixing virgin and recycled polymers [364]. This can also be achieved via traditional theoretical approaches to maximize the recycled polymer intake to obtain a given mechanical behavior [365]. Generally speaking, the development of theoretical models is able to predict the processing behavior as well as the final product properties that represents a crucial development [366]. Moreover, special attention is required in the general field of carbon footprint. LCA analyses involving the use of recycled polymers have been reported [367,368,369] and demonstrate that the electrical energy input is one of the main contributors to carbon footprint. Consequently, thermochemical process can be conveniently analyzed in this way despite the inherent complexity [370,371]. Many LCAs still miss the evaluation of the end-of-life of a given product [372] or are inherently focused on a specific waste/application area, e.g., windmills [369,373], electronics [374], asphalt [375], automotive [376], etc. We note the lack of a general LCA strategy across multiple recycling methods and application fields. These developments are crucial, especially when making allowances for the widespread use of LCA results in prioritizing investments and strategies as well as future research projects and government policies [377,378,379,380,381].

From an economic point of view, polymer recycling can be conveniently performed in an integrated manner by combining all recycling options (e.g., chemical and material) into a single facilities in order to minimize transport costs [382]. Moreover, significant advances in the recycling equipment design have also been described [383]. For example, by introducing nanoseconds, electromagnetic pulse-grinded waste polymers can be processed by decreasing the overall heat load, thus reducing the impact of thermal degradation [384].

A possible game changer in the recycling world is the rising use of electrochemical techniques. The general strategy is the one of oxidation (for example, with polyolefins [385], as shown in Figure 12).

The obtained chemicals can be re-used for the synthesis of polyesters and polyamides. Besides the fact that the obtained polymers are fairly different from the original polyolefins, the technique can also be used to depolymerize polyesters such as PET back to their original monomers [385].

Furthermore, from a technological viewpoint, recent developments can be mentioned as specifically contributing to material recycling. For example, extruders that are able to predict (based on the material properties as well as on the details of the extrusion process) and compensate for the change in viscosity along the barrel by adjusting the screw speed have recently been described [386,387]. Moreover, computational fluid dynamic (CFD) modeling of the extrusion process as well as of ancillary steps (e.g., melt filtration [388]) clearly and more easily render a mathematical description for subsequent upscaling.

Finally, an integrated approach, following some specific literature examples [389], should be used in order to tackle the challenges posed by recycling in a comprehensive manner. This conceptual integration nicely dovetails the more logistic one (vide supra). This is conceptually counterbalancing the fragmentation factors in industrial and consumer applications [390], which in turn dovetails the kaleidoscopic varieties of recycling strategies.

4. Conclusions

This work presents, in a compact format, a comprehensive overview of the state of the art in the general field of polymer recycling with specific focus on commodity polymers. This constitutes a relevant topic due to the increasing plastic consumption on one side and the depletion of fossil sources on the other. Against this backdrop, waste recycling, as a future carbon source, is gaining full attention at the academic and industrial levels, seen also as the worldwide commitment to become fully circular within the next 30–50 years.

Material recycling for thermoplastics is still investigated to reduce unwanted side reactions; for example, the degradation of polymers (e.g., PP, PVC) still hinders the recovery of the virgin material properties. On the other hand, relevant strategies can be identified in handling mixed polymeric waste, among which blending is the most popular. For polycondensates, the chemical breakage of the chain to the level of oligomers and/or monomers seems to be a viable route by combining relative high purities of the recycled materials with a significant retention of mechanical and rheological properties. Some of these chemical routes are also applicable for rubbers and thermosets where the necessity to break the bond between the chains is a conditio-sine-qua-non for material recycling. As numerous attempts in this direction on currently used materials (e.g., rubber devulcanization) have only reached relatively low TRL levels, chemical recycling strategies, at the moment, constitute the most popular choice.

In Future Outlook, we highlight more fundamental strategies (among others covalent adaptable networks) aimed at the design of novel polymeric products that possess inherent characteristics for recycling. At the same time, relatively new developments in electrochemistry, recycling technologies and the upgrading of recycled polymers via chemical modification offer many opportunities when dealing with current materials.

On a more general level, we note the fragmented nature of the research in this field, with defined expertise often focusing on a specific kind of waste, application fields and used technologies (e.g., material vs. chemical recycling). In contrast, the hypothetical journey of a single carbon atom might consist of different steps requiring different scientific insights and technologies. It might start from raw materials to monomer (e.g., butadiene) and then polymer in a daily product (e.g., in a tire). Its recycling might continue, for example, first as a part of new tire and then to a series of less valuable products (such as rubber foils for roof protection) up to a monomer again after chemical recycling (e.g., pyrolysis) and then having a new life as fuel or as a monomer again (thus re-starting the chain again). Understating and being able to control this journey requires a comprehensive approach on a scientific and technological level, but also, and perhaps crucially here, across different disciplines including environmental impact assessment and economic feasibility. As stated in the review (vide supra), a comprehensive and multidisciplinary approach is factually needed to be able to tackle the “recycling thematic” in a proper way.