Exploring Odor Minimization in Post-Consumer Plastic Packaging Waste through the Use of Probiotic Bacteria

: Plastic packaging represents a large proportion of the plastic consumption throughout the world. The negative environmental impact associated with plastic packaging waste can be in part abated by recycling plastics, and increasing numbers of regulatory frameworks are being adopted towards this goal. Despite recent advances in modern recycling technologies, the production of high-quality polyoleﬁn recyclates remains a challenge. Among other functional requirements, odor plays a crucial role in the acceptance of recycled packaging. This presents a challenge, as odor contamination in plastic packaging waste can stem from diverse sources, such as spoilage processes, and strongly depends on the quality of the post-consumer input material. The present study addressed this issue by exploring potential odor abatement of malodors in packaging waste through the use of probiotic bacteria. Speciﬁcally, probiotics were added to a mixed post-consumer plastic packaging waste fraction, which was subsequently evaluated using human sensory and gas chromatography–olfactometric analyses. A comparison of treated with untreated plastic waste fractions revealed signiﬁcant sensory di ﬀ erences. Further structural elucidation of the causative odorants conﬁrmed a reduction in malodorous microbial metabolites, although complete odor removal was not achieved. However, this environmentally friendly approach may represent an essential step towards overcoming the odor burden in post-consumer plastic packaging recyclates.


Introduction
Plastic is the number one material for packaging, as is evident from the vast amount of plastic packaging material produced globally every year, estimated at 78 million tons [1]. Despite these quantities, only a fraction is recycled, with 10% of the global plastic packaging flow currently part of a closed-loop (2%) or cascaded recycling (8%), while the majority enters landfill or is leaked into the environment [1]. To cope with growing quantities of plastic waste and to conform with increasingly stringent legal regulations for recycling quotas [2], a functional plastics recycling chain that yields high-quality recycled plastic material is more important than ever [3]. In relation to polyolefins, insufficient quality and, in particular, odor-related problems are still major impediments for the widespread re-use of recycled plastic materials [4]. The unwanted smell of recyclates can be traced back to the high contamination of input material, primarily from post-consumer plastic waste [5]. Although methods that aim at odor reduction exist, for instance by physical deodorization or advanced washing processes, including the application of hot water or chemical additives, a complete odor removal has hitherto not been achieved [6][7][8][9][10][11][12][13]. Moreover, a high-energy input or the addition of potentially environmentally critical chemicals, such as detergents, absorbers or entrainers, are often still required.
Ecologically friendly alternatives are in rising demand in many industrial fields. Probiotics, as valuable non-pathogenic bacterial cultures, are increasingly applied in many sectors, with the fields of human health and food quality representing the most traditional and best established areas [14,15]. The benefits of probiotics are manifold, including their potential ability to displace other microorganisms that may be harmful. This is achieved not only by the competition for nutrients and space, but also by their potential excretion of enzymes and antibiotic compounds that keep pathogens at bay [16]. More recently, the implementation of probiotics in livestock farming has been increasingly explored, especially in aquaculture [17][18][19][20]. A successful reduction of pathogens was even reported in hospital hygiene control using surface cleaning agents containing spores of diverse Bacillus spores [21]. Many probiotic strains additionally exhibit a potential to reduce organic matter via exo-enzymatic metabolism. Such reactions have been reported to support the degradation of biofilms, as well as the conversion of odorous contaminants into non-odorous volatiles, yielding innovative and promising areas of application in wastewater or industrial cooling waters [22][23][24][25][26]. For instance, a substantial reduction of a sulfurous, rotten egg-like smell was achieved by applying different bacteria of the genus Bacillus to industrial wastewater [27].
With regard to the smell associated with recycled plastics, the elimination or reduction of odor-active contaminants, or even avoidance of their formation at an early stage, is of prime importance. Recently, it was demonstrated that despite an observed decrease in odorous volatiles in mixed plastic film waste through washing, most of these components were still present after extrusion [28]. In terms of the extended storage periods of post-consumer plastic packaging waste, commonly up to three to four weeks in households, with several additional weeks until final processing, an early avoidance of spoilage processes presents a promising approach to reduce generation of odorous compounds from rotting processes of organic matter in plastic packaging waste [5].
In view of these challenges, the present study aimed not only at a molecular characterization of the causative odorous contaminants, but also at a reduction of odorous microbial metabolites in post-consumer plastic packaging waste by the application of a probiotic bacterial solution during the recycling process. Our focus was on the film fraction, primarily lightweight packaging, which has been previously reported to be especially susceptible to smell contamination [29]. To this aim, the efficiency of odor removal or even avoidance of odor generation in post-consumer plastic packaging waste using a commercial probiotics formulation was investigated on a pilot scale. Molecular characterization of the causative odorants in the probiotics-treated sample in comparison to an untreated reference sample was carried out using a combination of human sensory evaluations together with targeted high-resolution gas chromatography-mass spectrometry analyses.

Post-Consumer Plastic Packaging Waste Fraction
Post-consumer plastic packaging waste was sourced from household waste collected through a local council collection system known as 'yellow bag' collection in the state of Bavaria, Germany. The examined waste comprised the fraction no. 352 'lightweight packaging (LWP): mixed plastics' [30], that was sorted by means of state-of-the-art sorting technology (sorting date: 24 July 2019, Bavaria, Germany) [29] and chosen because of the typically intense off-odor associated with this post-consumer packaging waste. This fraction consisted mainly of plastic films that had been shredded post-sorting to 50 mm particles for storage. The waste contained 55% polyolefin and had a dry matter content of 93% (w/w).

Probiotic Bacterial Solution
A commercial bacterial suspension, 'PIP Aquatec Basic' (Chrisal), was applied, which is sold as a strong probiotic concentrate for the treatment and maintenance of large industrial water systems, such as cooling water and process water [31,32]. According to the manufacturer, the probiotic bacteria metabolize organic dirt and thus prevent water cloudiness, odor and biofilm formation. Further, it is claimed that the bacteria potentially displace microorganisms in water systems that produce odorous volatile gases while releasing odorless volatile compounds only. These positive effects are purportedly achieved by several different probiotic species (including different Bacillus spores) with 50 million germs per mL. For the present study, and according to the manufacturer's dosage, 2 L/m 3 was used for the treatment of the waste, which was diluted with water at a ratio of 1:10 for use as a spray solution. Appropriate adaption and growth of the probiotic bacteria on the input waste material has been ensured in microbiological pretests on a laboratory scale. Since the time of application was found to play an important role, probiotics were applied directly after sorting and shredding of the input waste material (Section 2.2.).

Pretreatment With Probiotic Bacteria and Washing Process of the Separated Post-Consumer Packaging Waste
Two 250 kg batches of post-consumer plastic packaging waste (fraction no. 352, cf. Section 2.1.) were sourced from a local sorting plant (date: 24 July 2019, Bavaria, Germany). Each batch was collected in a separate big bag (1 m 3 ). The first batch (250 kg) was sprayed with 20 L of the diluted probiotic bacterial solution; the second batch remained untreated and served as a reference. In order to ensure sufficient wetting of the input material and that the bacterial solution adhered evenly to the waste, 50 kg aliquots were placed on a conveyor belt and sprayed with 5 L of the bacterial solution. Afterwards, the bag was turned several times. This spraying and mixing procedure was repeated four times in total. A standard pressure spray bottle (Primex 5 Comfort, Gloria, Witten, Germany) fitted with a spray lance and nozzle with a barrel volume of 5 L was used to apply the bacterial solution. Both batches were subsequently stored outside for 40 days (24 July 2019-2 September 2019 Germany) until further processing.
A subsequent washing process after 40 days storage was carried out at a technical center (Herbold Meckesheim GmbH, Meckesheim, Germany). To eliminate any carry-over effects from water circulation (120 m 3 ), the circulated washing water was replaced with fresh water between each washing step of the individual samples. Both batches (sprayed waste and untreated reference waste) were cleaned with cold water in a two-stage washing process ( Figure 1). After the initial removal of coarse impurities in the pre-washing stage, a friction washer performed the main washing. The residence time of each waste input material within the wet processing was approximately 10 min. Subsequently, the sample material was first mechanically dried in a centrifuged dryer followed by thermal drying with a hot air flow of approximately 80 • C (T1015 Mechanical Dryer and TNT 500 Thermal Dryer, both Herbold Meckesheim GmbH, Meckesheim, Germany). Afterwards, aliquots of both waste samples were stored at −20 • C until sensory analyses to prevent any further bacterial activity.

Triangle Test
A triangle test according to DIN EN ISO 4120:2007 was performed to determine possible sensory differences between the samples; this forced-choice approach is applicable especially in such cases when the nature of the potential difference is unknown. Overall, 19 participants were asked to determine the deviating sample out of a series of three samples or, if no difference was perceivable, to randomly choose one of them. In each case, 2g (±0.1 g) of the reference (untreated) plastic packaging waste or the probiotics-treated waste was respectively placed into a 140 mL covered glass vessel, randomly distributed, encoded and presented to the panelists. This procedure was performed in triplicate so that each panelist had to evaluate three series (in total nine single samples), leading to a total of 57 responses. Based on a significance level of α = 0.001, at least 30 correct answers were required to prove a perceivable sensory difference.

Odor Profile Analysis
The in-house odor profile analysis was based on the industry standard DIN EN ISO 13299:2016-09. Panelists (cf. Section 2.3.1.) were required to describe their perceived orthonasal impressions, which were subsequently collated across the panel and shortlisted based on a consensus approval of the panel (more than 50% of panelists) to establish a selection of odor attributes that best characterize the perceived odor of the samples. Subsequently, each panelist was required to rate the intensity of each attribute in both samples on a scale from 0 (no perception) to 10 (strong perception), which was performed in comparison to odorant references that were presented in form of odorant pens. The reference compounds were 2-methylisoborneol (moldy/musty), geosmin (earthy), 3-phenylpropanoic acid (flowery/beeswax-like), (E)-2-nonenal (fatty/cardboard-like/cucumber-like) and octanal (soapy/citrus-like). Statistical differences were calculated by a paired, two-tailed distributed Student's t-test at α = 0.05.

Solvent Extraction of Volatiles
Solvent extraction proceeded immersing 10 g (±0.1 g) of the respective sample material in 200 mL of dichloromethane and stirring for 30min at room temperature. After filtration through cotton wool, the volatile fraction, including odor-active compounds, was gently separated by solvent-assisted flavor evaporation (SAFE) according to Engel et al. [33]. The SAFE procedure was carried out under high vacuum (10 −4 mbar) with a water bath temperature of 50 • C while the apparatus temperature was kept at 55 • C. After drying over anhydrous sodium sulfate, the distillate was concentrated to a total volume of~100 µL by means of Vigreux distillation and subsequent microdistillation [34]. The successful extraction of odor-active compounds was verified by comparing the odor impression of a single drop of the obtained distillate applied on a filter paper with the odor of the respective input waste material. Details of chemicals and reference compounds used are listed in Appendix A.

Comparative Odor Extract Dilution Analysis
Relative odor intensities of single odor-active regions in gas chromatography-olfactometry (GC-O) (cf. Section 2.4.3) between treated and untreated samples were compared via comparative odor extract dilution analyses (cOEDA) [35,36]. The undiluted distillates (cf. Section 2.4.1), corresponding to odor dilution (OD) 1, were diluted stepwise (1 + 2 v/v) with dichloromethane to produce solutions for GC-O analysis that corresponded to OD factors 3, 9, 27, 81, 243 and 729. The higher the OD factor of a detected compound, the more relevant it is for the overall odor of the sample.

Gas Chromatography-Olfactometry
The undiluted distillates and 2 µL aliquots of each dilution (cf. Section 2.4.2) were analyzed by GC-O with a Trace GC Ultra (Thermo Fisher Scientific GmbH, Dreieich, Germany) using two capillary columns: DB-FFAP or DB-5 (both 30 m × 0.3 mm, film thickness 0.25 µm; J&W Scientific, Agilent Technologies, Waldbronn, Germany). Sample injection was applied using the cold on-columntechnique with an initial oven temperature of 40 • C, which was then raised at 8 • C/min to either 235 • C (DB-FFAP) or 250 • C (DB-5). The final temperature was held for 5 min in both cases. Helium was used as carrier gas at a flow rate of 2.2 mL/min. At the end of the capillary column, the effluent was equally split in two parts and transferred via a Y-type glass splitter and deactivated fused silica capillaries (0.5 m × 0.2 mm, A-Z Analytik-Zubehör GmbH, Langen, Germany) to a flame ionization detector (FID, 250 • C) and an odor detection port (ODP, 270 • C). GC-O analyses of the original distillates were performed by three trained panelists.
The tentative identification of odorants was based on linear retention indices (RIs) on the two capillary columns of different polarities (DB-FFAP and DB-5) and calculated by analyzing a homologous series of n-alkanes [37], together with the perceived odor quality at the ODP in comparison to the respective reference compound. For further unequivocal identification, the molecular structure was elucidated by mass spectral data (if obtained) and also compared to reference compounds (cf. Section 2.4.4).

Two-Dimensional Gas Chromatography-Mass Spectrometry/Olfactometry
Two-dimensional gas chromatography-mass spectrometry/olfactometry (2D-GC-MS/O) was performed for the unequivocal identification of odorants based on mass spectral data using a system of two gas chromatographs (Varian CP-3800, Agilent Technologies, Waldbronn, Germany) equipped with a DB-FFAP column in the first dimension and a DB-5 column in the second dimension (cf. Section 2.4.3. for capillary column parameters). Both GCs were connected via a CTS 1 cryotrap system cooled with liquid nitrogen to −100 • C (Gerstel GmbH & Co. KG, Mülheim, Germany). Automated sample injection of 4 µL of the original distillates (OD1) by the cold on-column technique was performed with a MPS 2XL multipurpose sampler (Gerstel). After 2 min at an initial oven temperature of 40 • C in the first GC system, the temperature was raised at 8 • C/min to 230 • C and held for 5 min. The helium carrier gas flow was 9.0 mL/min. At the end of the first GC system, the effluent was split (1:1) via a Y-type glass splitter to an FID (250 • C) and an OPD (270 • C).
The area containing the analyte of interest was separated after passing the first GC system and transferred to the cryotrap by means of a MCS2 multi-column switching system (Gerstel). The subsequent transfer of volatiles to the second GC system was achieved by thermodesorption at 250 • C. In the second GC system, the initial temperature of 40 • C (2 min) was raised (8 • C/min) to 250 • C (1 min), with the effluent at the end of the capillary column again split in two equal parts, allowing the simultaneous detection of odor qualities at the ODP (270 • C) and mass spectra by means of a Saturn 2200 ion trap mass spectrometer (Agilent Technologies, Waldbronn, Germany). Mass spectral data were generated through electron ionization in positive mode at 70 eV ionization energy, over a m/z range of 35-399.

Sensory Evaluation
With the initial focus on revealing possible odor differences, the post-consumer plastic packaging waste sample that had been treated with the probiotic bacterial solution was compared to the untreated reference sample via a triangle test (cf. Section 2.3.2). This forced-choice test returned 34 correct responses from a total of 57, indicating that the panel was able to discriminate the samples based on the perceived odor (α-risk: 0.001).
To further characterize the odor difference, 14 trained panelists agreed on five odor attributes as descriptors for both samples (cf. Section 2.3.3). These included moldy/musty, earthy, flowery/beeswaxlike, fatty/cardboard-like/cucumber-like and soapy/citrus-like. The highest intensity ratings in both samples (treated and untreated) were for the attributes moldy/musty (mean intensity ratings of 6.7 and 7.3) and earthy (mean intensity ratings of 5.4 and 5.6), with ratings similar for both samples (Figure 2). By comparison, the attributes fatty/cardboard-like/cucumber-like and soapy/citrus-like were rated with lower but comparable intensities of 3.1/2.6 and 3.1/2.9 (reference sample/probiotics-treated sample). The lowest intensities were obtained for the attribute flowery/beeswax-like, which was rated at an intensity of 1.4 in both samples. Despite the differences perceived between the samples in the triangle tests, the mean intensity rating differences of individual attributes between the samples during odor profile analysis were not significant (paired, two-tailed distributed Student's t-test at α = 0.05). This might be attributable to the relatively high variance in individual intensity ratings for single descriptors, as is regularly observed in sensory evaluations, possibly due to differences in individual odor sensitivities and/or odor thresholds [38].

Identification and Characterization of Causative Odorants
Gas chromatography-olfactometry (GC-O) analyses on the original distillate of the untreated reference sample performed by three trained panelists led to the detection of 57 odor-active regions ( Table 1). Further analyses of each dilution (cOEDA, cf. Section 2.4.2) allowed an assignment of the most potent odorants to the respective odor dilution (OD) factors.
Of the total 57 odorous contaminants detected in the untreated reference sample, 43 substances were also perceived in the undiluted distillate (OD 1) of the plastic packaging waste sample that had been treated with the probiotic bacterial cultures. On the other hand, no additional odorants were detectable, and only four odorants, namely α-isomethylionone (36), verdyl acetate (38), the unknown mushroom-like smelling compound (54) and γ-decalactone (48), were identified in the treated sample at the highest OD factors of 729 or 243, respectively. A direct comparison of both samples further revealed that 15 of the 43 odorants were detected at the same OD factor in each case. In contrast, and with the sole exception of 1-hexen-3-one (3), the remaining 27 odorants showed lower OD factors in the treated sample, indicating that more than 70% of the odorous contaminants were perceived at lower intensity during cOEDA. Odorous compounds that differed by more than three dilution steps between the two samples were 2-mercapto-3-pentanone (12), 2-methylisoborneol (23) and p-cresol (47).

Comparison of Human Sensory Evaluation and Analytical Results
The triangle test yielded significant differences in the overall odors of the untreated reference sample and the sample treated with probiotic bacterial cultures. To further objectify this difference an odor profile analysis was performed, which confirmed a major odor load for both samples. Overall, the moldy/musty and earthy smell impressions dominated the overall profiles of both samples, with both attributes rated with high intensities (6.7/7.3 and 5.4/5.6, respectively). Comparison with the GC-O data indicated that 1-octen-3-one (8), trimethylpyrazine (13), octanoic acid (45), patchouli alcohol (49) and especially the earthy/musty smelling 2-methylisoborneol (23) are the main compounds likely to be responsible for this odor, especially the latter, which yielded the highest OD factor in the reference sample. Further, the fatty smelling odorants (E)-2-octenal (14), (E)-2-nonenal (20) and (E,E)-2,4-nonadienal (28) are likely to contribute to the fatty odor impression rated with moderate intensities in both samples (3.1/2.6). Interestingly, the attributes soapy and flowery were rated with the lowest intensities despite the fact that the instrumental olfactometric analyses revealed a variety of flowery and soapy smelling substances (no. 7, 10, 21,34,36,39,43). While most of these flowery or soapy smelling compounds were perceivable at the highest OD for both samples, they are likely to be covered by other potent malodorous compounds, for example the moldy smelling 2-methylisoborneol. One reason might be that complex mixtures of odor-active compounds may cover or enhance the impact of single odorants, even to an extent that individual odor impressions are no longer attributable to single odorous compounds [35,[39][40][41]. It should also be noted that many of these flowery and soapy smelling substances exhibit a higher volatility. Due to the complete evaporation of the distillate on the GC column compared to vapor pressures at room temperature, higher OD factors are therefore less important for these odorants [42].
The complexity of the odorant composition might also explain why the two samples did not exhibit significant differences in any of the rated attributes despite the clear discernibility of the samples in the triangle test. Additionally, as determined by cOEDA, potent odorants (OD factor ≥ 27 in the reference sample) with the highest differences in OD factors between the reference and the probiotics-treated sample (no. 9,11,12,17,23,25,37,45,47) showed deviations of a maximum of three OD steps. However, these compounds were still perceivable in both samples in any case, meaning that none of these compounds was fully eliminated through the probiotic treatment. This might explain why the odor profile analysis did not provide a satisfactory differentiation between the two samples on the basis of the defined odor attributes. Otherwise, in the case of the attributes soapy and flowery, which were rated with similar intensities in both samples, the sensory evaluation correlated well with the cOEDA results, as most flowery or soapy smelling substances were detected at the same OD factor in both samples. Accordingly, cOEDA served as a necessary approach to distinguish molecular differences in single odorous constituents between the samples.

Potential Origin of Identified Odorants
A large number of odorous contaminants was identified in the post-consumer plastic packaging waste of the present study, which is in agreement with previous studies on household plastic waste or recycled plastic materials [5,9,29,40,43]. However, 17 of the detected odorants (no. 2, 3, 6, 10, 13,19,30,31,33,34,37,40,41,45,46,52,57), which represent around 30% of all detected substances, are reported here for the first time as odor-active contaminants in plastic packaging waste and recycled plastics. A reason for this large number of previously unreported odorants might be the inhomogeneity of post-consumer material, and variations in degree of contamination and, accordingly, odor load.
Generally, odorants stem from various sources or via diverse formation pathways. The majority of the odor-active compounds identified here represent typical metabolites of microorganisms, such as 2-methylisoborneol (23), p-cresol (47) and diverse short-chain carboxylic acids (cf. Section 4.3). This indicates that the off-odor primarily originates from substances emitted by bioconversion of organic residues typically found in post-consumer waste. Only about 20% of the identified odorants most likely stem from residual fragranced filling goods, such as washing and cleaning agents. Compounds such as α-isomethylionone (36) and verdyl acetate (38) are especially commonly used as fragrance compounds [44]. The pronounced appearance of microbial metabolites is in line with previous findings on post-consumer plastic packaging waste and plastic film fractions [5,29,45]. Deviant from the smell composition of the mixed polyolefin fractions, recycled post-consumer high-density polyethylene (HDPE) and polypropylene (PP) has been found to be additionally dominated by the soapy/flowery smells, corresponding to the detection of typical fragrance compounds, as discussed above [9,13,40].
Additionally, several odorants are also likely to stem from the plastic polymer, and possibly additives, representing oxidation and degradation products thereof. In previous studies, unsaturated aldehydes, short-chain carboxylic acids and γ-lactones have been reported as odor-active compounds in plastics [46][47][48][49].

Odor Reduction by the Application of Probiotic Bacterial Cultures
Significant differences between the odors of the untreated reference sample and the sample treated with probiotic bacterial cultures were found via triangle tests and cOEDA. Thereby, ten odorants exhibited lower OD factors by two or more dilution steps, and 14 odorous contaminants present in the reference sample were not perceived at all in the undiluted distillate of the probiotics-treated sample. Although cOEDA is only a screening method to obtain an approximate odor contribution weighting of the individual substances, it nevertheless serves the goal of obtaining an indication of relative quantitative differences [36]. Accordingly, major differences in OD steps or even non-detection versus detection between samples is a strong indicator that the related odorant quantities were influenced to a major extent by the process under investigation.
The vast majority of odorants showing noteworthy OD differences in this study most likely relate to the bioconversion of organic residue materials, since they have been reported elsewhere to be common metabolites of microorganisms. In particular, especially the group of short-chain carboxylic acids (no. 16,25,29,41,45) and methylated carboxylic acids (no. 22,27,31,33) exhibited lower OD factors in the probiotics-treated sample or were not detected at all. Generation of volatile acids has been reported for different bacterial species, such as Clostridium, Lactobacillus, Pediococcus and Leuconostoc [50,51], and might explain their previous detection in organic waste and wastewater [52][53][54]. Various other compounds can also be categorized as bacterial metabolites of fatty acids, for instance 2,3-diketones (no. 1), diverse esters (no. 2, 4) and lactones (no. 44,48,51,53), which have been reported to be generated by microorganisms such as Staphylococcus, Stigmatella, Loktanella and Dinoroseobacter [55,56]. On the other hand, such low molecular weight compounds themselves may serve as nutrient and energy source for the microbiota. In the present study, the majority of the detected fatty acid derivatives was perceived at lower OD factors in the probiotics-treated sample than the untreated reference sample, indicating a notable trend of reduction by the application of the probiotic bacterial solution.
Unlike the detected microbial metabolites, the majority of typical fragrance constituents (no. 7, 10, 21,32,34,36,38,39,42,49) was perceived at the same OD factors in both samples (Figure 3). Especially α-isomethylionone (36) and verdyl acetate (38) were perceived even into the highest dilution of the two samples by means of olfactory detection, substantiating the comparable contamination of both samples with regard to typical fragrances. Consequently, the present study shows that, in this case and under these process conditions, probiotics are not capable of degrading such fragrance-related odorants. As such, the contamination degree of the input plastic packaging waste material with these fragrances remained unaffected by probiotic bacterial treatment. This finding calls for new strategies in the fragrance industry in view of designing recycling and biodegradation strategies for filling goods and scent constituents, appropriate measures to avoid fragrance compound contamination of the plastic material, and additional decontamination strategies, ideally combined as a systemic strategy. The reduction in OD factors between the untreated and treated sample indicates a quantitative reduction of these odor-active compounds in the plastic packaging waste fraction after probiotic treatment. Moreover, no additional odorants were detected in the latter, indicating that only non-odorous derivatives were formed, while odor-forming microorganisms may have been additionally limited in growth. Nevertheless, a complete inhibition of odor-forming microbial activity was not achieved, potentially due to the initial odor load being too abundant to be fully removed. Hence, the treated sample still exhibited a relatively strong odor.
On the contrary, the probiotic bacteria affecting other properties of the plastic material is unlikely given that the probiotic aqueous solution did not contain any further chemicals. Additional moisture content may not be of importance since the material was mechanically and thermally dried at the end of the evaluated recycling process. Regarding the applied probiotic bacteria themselves, a modification of plastic polymers is unlikely, given that in literature it was only shown for specifically isolated bacteria to be able to decompose polyethylene terephthalate (PET) [65]. On top, common post-consumer plastic waste also contains various kinds of probiotic bacteria, originating, amongst others, from diverse food residues such as dairy products, whilst a modification of the plastic packaging has not been reported so far.
In short, we deem the implementation of probiotics in the plastics recycling chain as a promising strategy towards the reduction of unwanted organic matter and odor. It is important to mention that post-consumer plastic packaging waste is commonly stored for extended periods of up to several weeks or months starting from the collection in households until usage for recyclate production. This favors abundant spoilage processes and, consequently, odor contamination. New strategies in the collection, logistics and storage of plastic packaging waste, together with the right timing of implementing treatment with probiotics, are in our opinion crucial for successful implementation of such biological strategies into a systemic waste management and recycling concept. This is especially true when it comes to potent odorants that annoy consumers even at lowest concentration levels. Their removal will be essential for consumer acceptance and future recycling strategies for deodorizing post-consumer plastic packaging waste.