Effect of Different Natural Fibers on Mechanical and Disintegration Properties of Compostable Biobased Plastics

ABSTRACT This study evaluates the economic and degradation properties of natural fiber-reinforced PHBH (Poly(3-hydroxybutyrate-co-3-hydroxyhexanoate)). Compounds containing the compostable biobased plastic PHBH and 30 mass% of cotton, merino wool, wood fibers, and hemp fibers were produced, and 0.4 mm thick foils were manufactured using a laboratory hot press. Samples for mechanical characterization were produced by injection molding and were buried in bioactive soil for 16 weeks to determine the influence of fiber type on the disintegration speed in soil. The disintegration was measured by subsequent excavation, determination of dry mass loss and visual inspection. All kinds of natural fibers generally improve disintegration speed. The most significant improvement was reached by samples containing cotton and wool, respectively, which experienced similar mass loss, followed by the ones containing hemp fibers. The smallest increase in disintegration speed was achieved by samples containing wood fibers. Mechanical properties and cost-saving potential increased in the cotton and hemp samples. * Long version of a presentation at Composites Meet Sustainability – Proceedings of the 20th European Conference on Composite Materials, ECCM20. 26–30 June, 2022, Lausanne, Switzerland


Introduction
In the past few years, several papers have highlighted the growing concern about microplastics in the ocean and the distribution of microplastics in all oceans.They conclude the need for a global solution to reduce microplastic pollution and the potential harm to the ecosystem (e.g., Andrady 2017;Cole et al. 2011).Others report on the ingestion of microplastics by marine animals and summarize the evidence for the presence of microplastics in seafood and their potential effects on human health and aquatic organisms (see, e.g., Smith et al. 2018).Fractions of micropollutants can also be found in drinking water, as Koelmans et al. (2019) report.Recently, more and more studies have shown that microplastics are also widespread in the soil environment (e.g., Boots, Russell, and Green 2019;Guo et al. 2020;Scheurer and Bigalke 2018).
Compostable thermoplastics may solve severe littering and microplastic problems in a large range of products (Bauchmüller et al. 2021).While these polymers' production capacities and demands are rising, the price is still relatively high.In some cases, natural fibers can reduce material costs, but this only makes sense if the mechanical properties of the composite are not significantly worse than those of the unreinforced polymer.To assess the suitability of a material to substitute an existing one without sacrificing performance and at the same time reducing cost can be evaluated using material indices, depending on the mechanical load case (Ashby 2005).
Depending on the application, an increased rate of material disintegration may be beneficial.Due to the chemical compositions of different types of natural fibers, the degradability in soil differs significantly.Wool seems to be particularly interesting in this context.Röhl and Müssig (2022) reviewed the current research and development in wool fiber-reinforced thermoplastic composites.They aimed to demonstrate the capabilities of composites containing keratin fibers and to emphasize the unique characteristics that must be considered when developing and producing wool fiber-reinforced composites.
In developing fiber-reinforced composites, the classical approach is to insert a stiff fiber into a compliant matrix.Hemp and flax fiber-reinforced composites (NFRC) have proven successful.However, hair fibers like wool have lower strength and stiffness than bast fibers such as flax or hemp (Bourmaud et al. 2018;Müssig et al. 2020).The tensile behavior of wool fibers is different and shows a yielding behavior dominated by molecular unfolding, resulting in a tensile curve with three different stiffness ranges and elongations at break of over 30 %.The macro-and microstructure of wool fibers are crucial for understanding their mechanics.While wool fibers have a Young's modulus of about 3000 to 5000 MPa, the final tensile strength is only reached at elongations above 30 %.All this makes wool fibers an ineffective reinforcement when mixed with low elongation at break matrices (Röhl and Müssig 2022).
Using natural fibers as structural reinforcement is challenging because animal hair can only reach a limited length (see, i.e., Hunter 2020).The hydrophilic cortex and hydrophobic cuticle of wool fibers lead to their unique relationship with water, which is challenging when using wool as a structural reinforcement.The fiber-matrix interaction is essential for transferring stress and depends on the fiber surface, chemical compatibility and mechanical interlock (see, i.e., Müssig and Graupner 2020).To reduce the critical fiber length, either the fiber diameter must be reduced, or the strength of the fibermatrix interface must be improved (Kelly and Tyson1965).Optimization of fiber-matrix interaction in NFRCs containing plant fibers in thermoplastic matrices has successfully incorporated bast fibers into hydrophobic matrices such as polypropylene (Arbelaiz et al. 2005;Müssig and Graupner 2020).Röhl and Müssig (2022) critically evaluated and classified published work in wool fiber-reinforced thermoplastics according to thermoplastic material and processing technique.The properties obtained as a function of the processing methods are compared and used to describe the micromechanics and fiber-matrix interaction (see Figure 1).
In animal hair, the disulfide bond between cysteine molecules can be broken by keratinolytic microorganisms under the appropriate environmental conditions, and the hair can be metabolized (Solazzo et al. 2013).During the biodegradation of wool in soil, sulfur and nitrogen are made available to the plants and act as fertilizers.Górecki and Górecki (2010) have shown that 1 g of washed wool per cubic decimeter of soil can increase the fruit yield of tomato and pepper plants by up to 30 %.
When searching for suitable polymer materials for metabolizable composites, the choice quickly turns to microbial reserve polyesters such as polyhydroxyalkanoates (PHA) or poly(3-hydroxybutyrate) (PHB) (see, e.g., Anderson and Dawes 1990;Mergaert et al. 1992;Tokiwa and Calabia 2004).The following aspects are particularly worth exploring: (1) Approaches can be found in the literature to mix PHAs with natural fibers or fillers to manufacture a composite material reducing the compound's high polymer costs (see, e.g., Barkoula, Garkhail, and Peijs 2010;Sabapathy et al. 2021).However, there are usually no approaches to evaluate the costs concerning the material properties.(2) Furthermore, there are partly contradictory results on the degradability of PHA composites as a function of the fillers or natural fibers used, for example, Sabapathy et al. (2021) and Carofiglio et al. (2017).
Our study combines these two aspects and uses natural fibers with different potentials for soil disintegration, a large span of material costs, and widely differing properties.The aim is to develop an approach for the economic evaluation of the material property and to conduct a better monetary comparison of materials.Conducting a soil burial degradation test with different natural fibers processed into the same matrix makes sense.In this context, the question arose of how the disintegration rate in bioactive soil of the compostable biobased plastic PHBH changes by adding different natural fibers.

Materials used
Compounds were created according to the recipes in Table 1 using different natural fibers, including bast fibers, wood fibers, animal hair and seed fibers.PHBH-type Kaneka X331N (density: 1.20 g/cm 3 ; melting temperature: 145 °C; T G : 2 °C) was supplied as a polymer matrix by H&P Moulding Emmen B.V. (Emmen, NL).This polymer's market price at that time was roughly 10 €/kg.Polysaccharide α − 1,3-glucan (PS), Nuvolve TM , an engineered polysaccharide polymer product from DuPont, BioMaterials, Wilmington, USA, was used for one compound (Ganesarajan et al. 2022).We have used the same price as for PHBH polymer for simplification.The prices and sources for the fibers are listed in Table 1.Prices for Merino wool at the given fineness are taken from the Micron Price Guide (MPG 2022) for Week 35 in 2021.The cotton price is assumed, given the current market reports and currency rates (USDA 2022).

Compound manufacturing & composite production
The fibers (30 mass%) were compounded with the matrices mentioned in Table 1 with a corotating CMS-compounder at 3N, Werlte, Germany (Müssig et al. 2020).Subsequently, injection molding (Emac50, ENGEL AUSTRIA GmbH, Schwertberg, A) was performed: screw temperatures were set to 150 °C (zone 1), 165 °C (zone 2) and 175 °C (zone 3); nozzle @ 175 °C; mold @ 25 °C.The maximum pressure was set to 1200 bar, the injection speed was 24 cm 3 /s, and the cooling time was 30 s; the total cooling time was 55 s, including the pressure holding time.The holding pressure was 650 bar for 18 s and then gradually released for 7 seconds.
For the disintegration testing, foils of thickness 0.4 mm were produced using a laboratory hot press (LaboPress P200S, Vogt Labormaschinen GmbH, Berlin, DE).The pressing temperature was set to 170 °C at a pressing time of 5 minutes.Before pressing, the 10 g of the granules, previously dried for a minimum of 24 hours at a temperature of 50 °C, were placed in the press and given 5 minutes to warm up.The final thickness was achieved using Teflon foil as spacers.Eighteen strips of size 2 × 8 cm 2 were cut from each sample, so three samples could be dug up at each sampling point (after 1, 2, 4, 8, 12 and 16 weeks).

Composite -physical testing
The sample's dry mass was measured, and afterward, the samples were buried into bioactive soil in an indoor testing facility at room temperature.The soil's moisture was checked weekly, and water was added if necessary.Mechanical properties were tested on seven injection-molded samples (type 1A; DIN EN ISO 527).Tensile testing was performed on a Zwick/Roell Z020 universal testing machine (Zwick GmbH and Co., Ulm, DE) equipped with a 20 kN load cell, using a testing speed of 2 mm/min.The strain was measured using a video extensometer between two measuring marks spaced 50 mm apart (VideoXtens, Zwick/Roell GmbH, Ulm, DE).The Charpy impact strength was done (DIN EN ISO 179) with a pendulum device (type 5101 (Zwick GmbH, Ulm, DE) using a 2 J impact hammer.Six unnotched samples with a length of 80 mm, width of 10 mm and thickness of 4 mm, cut from the tensile bars, were used for the analyses.

Composite -material indices for the cost-saving potential
A model was used to determine the cost for a plate under uniform load following Ashby (2005) to assess the cost-saving potential of the different compounds in applications limited by tensile strength performance.Material indices for the cost-saving potential were derived, considering raw material and manufacturing costs as well as mechanical properties.For this, a compounding cost of 0.70 €/kg was assumed.A higher index value resembles a greater cost-saving potential.The material index for a plate under uniform load is derived as: R m : Tensile strength; C v : Cost per unit volume The costs per unit volume consider the material density, prices of fiber and matrix, and compounding costs.

Composite -disintegration experiments
Bioactive soil was prepared by adding 10 % of active compost.The soil humidity was adjusted to 80 % of the maximum water-holding capacity.The soil's moisture level was kept constant, and the temperature was 20 °C.The dry mass of all samples was measured, and afterward, they were buried in the soil.After 1, 2, 4, 8, 12 and 16 weeks, three specimens were dug up and weighed for each compound.

Mechanical characterisation
Figure 2 shows the mean stress-strain curves from the tensile tests of the unreinforced and reinforced PHBH samples.How strongly a particular natural fiber's choice can influence the curve's habitus is visible.
Adding wood fibers leads to increased stiffness (increased Young's modulus) with embrittlement of the material compared to the pure matrix.The increase in impact strength is particularly noteworthy compared to the pure matrix by adding wool and cotton.The results are summarized in Table 2. Müssig et al. (2020) processed with the same compounding and injection molding device used in the study of this paper, hemp with PLA, and showed that the measured object lengths and widths of the different hemp varieties extracted from the injection molded samples were comparable.Some fiber showed length values above 2000 µm, while the median was around 255 µm.Cotton fibers were extracted from the matrix by dissolving the PHBH with chloroform.The average fiber length was 1.26 mm, with maximum fiber lengths of up to 3.9 mm.
The results of the tensile tests for the pure polymer, with Young's modulus of 2 GPa and tensile strength of 27 MPa, are comparable to published values in the literature (see, e.g., Barkoula, Garkhail, and Peijs 2010).Young's modulus was improved in all samples compared to the neat matrix.The highest stiffness was reached by using hemp fibers, with cotton being close below at roughly 6 GPa.This is quite surprising, as common literature usually gives a Young's modulus of 10-12 GPa for cotton fibers.When modeling the expected composite Young's modulus using the Halpin-Tsai model for 2D-randomly oriented short fibers (Tsai and Kardos 1976), the results, even assuming very optimistic fiber lengths of 2 mm, are nowhere near the measured values.Concerning tensile strength, the cotton samples reach the highest values, with an increase of over 50 % compared to the neat matrix.While hemp fibers also increased the tensile strength of the composites, even though not as severely as cotton fibers, the use of wool and wood fiber decreased the strength compared to the neat matrix.Sabapathy et al. (2021) showed that PHB powder extracted from Acinetobacter junii BP 25 reinforced with sugarcane bagasse fiber improved Young's modulus from 416 to 740 MPa (30 mass % fiber content).The tensile strength of the pure polymer matrix in the publication mentioned before was 13.5 MPa and was only slightly changed to a value of 14.2 MPa with the addition of 30 mass% fibers.Carofiglio et al. (2017) investigated using olive mill wastewater residue (OWMR) for PHB composite-based film.Adding 30 mass% OWMR leads to a decrease in properties for Young's modulus and tensile strength.Table 2. Results of the mechanical testing and the calculated material index.For materials with indices higher than the value of the neat matrix (0.439), the cost can be saved by substitution.

Nr.
Young's Modulus in GPa Tensile strength in MPa Impact strength in kJ/m 2 Density in g/cm 3 M 1 -Cotton 6.0 ± .242.6 ± .5 18.9 ± 3.8  Barkoula, Garkhail, and Peijs (2010) significantly increase Young's modulus for flax and PHB composites from approximately 3 GPa for pure PHA to over 7 GPa for 30 mass% flax in PHB.The strength, in contrast, drops from about 40 MPa for the pure PHB to about 33 MPa for 30 mass% flax.The flax fiber-reinforced composites of polyhydroxybutyrate (PHB) and 12 % hydroxyvalerate (HV) show lower values.The copolymer PHB/HV shows a Young's modulus of less than 2 GPa; with 30 mass% flax, the value increases to approx.4 GPa.The strength values of the two variants are constant at about 30 MPa.The results mentioned above from Barkoula, Garkhail, and Peijs (2010) agree pretty well with our investigations on hemp composites, although the adhesion between fiber and matrix must be better developed in our series of experiments, as we achieved an increase in strength.In the study by Gunning et al. (2013), adding 30 % hemp, jute, and lyocell fibers decreases the composite's strength.Smith et al. (2020) achieved an increase in Young's modulus but decreased tensile strength in PHB composites with 25 % agave fibers (Agave tequilana) by mass and attributed this to poor adhesion.
The properties of our achieved tensile strengths of natural fiber-reinforced PHBH can be evaluated as very successful.Our research shows that the careful compounding of the natural fibers led to a good composite material, even without coupling agents.Noteworthy is the high tensile properties achieved for cotton fiber-reinforced PHBH.Promising results were obtained in testing the impact strength, where cotton samples gained the highest values.The second highest values were reached by merino wool samples, even though tensile strength was reduced in these samples, which could be due to the high strain potential of these fibers.This effect was not visible in the samples with merino wool and the 20 % PS (Nuvolve TM ) in the matrix, which could be due to a reduced deformability of the matrix due to PS (Nuvolve TM ) presence.In all other samples, impact strength was reduced compared to the neat matrix.In our work, using a bast fiber like hemp leads to a slight decrease in impact strength for unnotched specimens.In the work of Barkoula, Garkhail, and Peijs (2010), a positive influence of using flax fiber on the impact properties of PHB or PHB/HVB was shown for notched samples.This seems surprising at first glance, as in the work of Barkoula, Garkhail, and Peijs (2010), the elongations at break of the samples decrease significantly with the addition of flax.An explanation for this apparently contradictory behavior can be found in the influence of the notching.In our results for the unnotched hemp specimens, the added fiber acts as a crack initiator.In Barkoula, Garkhail, and Peijs (2010), the notch eases crack initiation, and the crack can deviate at the inserted fibers; thus, more energy can be absorbed.

Economic consideration
Cost savings can be realized for the cotton and hemp samples when calculating the economic material indices.This is due to the improved tensile strength and the relatively low price for raw fibers.No cost savings can be directly achieved for the wood samples.But because there was only a slight difference in the material index, this material may still be used to replace the neat matrix if an increased tensile stiffness is desirable.Due to the high price of merino wool and low tensile strength, product costs cannot be reduced when using this material.This is to be expected as fine merino wool is sought after for use in garments and textiles.A possibility would be to use coarser wool types, often discarded and therefore low-priced.But as previous investigations have shown (Röhl and Müssig 2022), a drop in tensile strength is expected due to lower fiber-matrix interaction.

Disintegration tests
We observed significant disintegration in the soil for the pure polymer PHBH-type, Kaneka X331N, used in our study (see Figures 3 and 4).The pure polymer showed the lowest degradation rate compared to all composites.One of the desirable properties of PHAs is their residue-free environmental disintegration which occurs mainly through the enzyme activities of microorganisms (Anderson and Dawes 1990;Emadian, Onay, and Demirel 2017;Fernandes et al. 2020).In the case of poly-β-hydroxybutyric acid (PHBS)-storing bacteria, Chowdhury already investigated in 1963 how the released polymer is degraded in the soil when the bacteria die and which enzymes are involved in PHBS disintegration.For pure PHB and a copolymer PHB/HV, Mergaert et al. (1992) showed that biodisintegration took place in all the environments they investigated.The copolymer showed a tendency to degrade more quickly.Tokiwa and Calabia (2004) address the factors influencing PHB's microbial and enzymatic degradation and biodegradability.They compare the degradation of PHB with other polymers such as PLA and observe, for example, that a PHB film was degraded entirely after storage for 6 d at 50 °C (Streptomyces strain MG), while the PLA film showed no degradation at all.Gómez and Michel (2013) studied the biodegradability of different polymers and natural fiber-reinforced composites in different environments.The relative biodegradability of the  materials during long-term soil incubation trial was ranked as follows; from fast (1) to no degradation (7): 1. PHA, 2. co-polyester + corn-based plastic, 3. composted cow manure, 4. plastarch (blend of polypropylene with corn starch), 5. paper pulp, 6. natural fibers, and 7. conventional plastics with additives to improve biodegradation, same as conventional plastics.
Looking strictly at the results of Gómez & Michel (2013), a mixture of natural fibers and PHA could reduce the degradation rate.Sabapathy et al. (2021) observed the degradation of PHBbased composites with the reinforcement of sugar cane bagasse in a soil compost under laboratory conditions, over 20 weeks, with degradation exceeding 80 % within 18 weeks.The addition of the fibers slowed down the degradation rate.Carofiglio et al. (2017) reported opposite results for olive mill wastewater residues (OMWR) PHB composite films.An addition of 30 % OMWR resulted in a faster degradation rate than the unreinforced plasticized PHB.Robledo-Ortíz et al. (2021) provide a comprehensive overview of work in lignocellulosic materials as reinforcement of PHB and its copolymer with hydroxy valerate.However, they do not mention any work using fibers like cotton or wool.
In our study, using natural fiber composite films containing a PHBH polymer resulted in faster disintegration, which seems to be a function of the natural fiber used.The visual disintegration of the buried pieces is shown in Figure 3.There is a difference between the neat samples and the other samples after one week.One can already see some color changes and increased haziness of the samples.In the first week, a biofilm is formed at the samples' surface.Visually, the highest disintegration was present in the wool samples, with only small fragments left after 16 weeks.Cotton samples, followed by hemp, achieved the second-highest disintegration.Visually, the neat matrix and the samples containing wood fibers are only slightly damaged, with an intense color change and loss of transparency.
Figure 4 shows the relative mass loss of the samples over the course of 16 weeks.The colored areas show the standard deviation range inside each sample.It is visible that the largest mass loss is achieved for the specimens containing wool, followed by the ones containing cotton.The standard deviation of the mass loss after 16 weeks is very high in the specimens containing cotton and hemp; no significant difference can be stated.The visual inspection underlines that disintegration is more substantial in the wool specimens.The lowest mass loss was measured in the neat matrix, with the introduction of short wood fibers leading only to a slight increase in disintegration severity.When the samples started falling apart, most of the remaining material was removed from the soil.Possibly some parts were overseen or too small to take out, which increases the standard deviation.
Zaidi, Mawad, and Crosky (2019) observed a significant acceleration of the degradation rates in the neat PHBV system compared to flax fiber-reinforced PHBV.Accelerated aging (cyclic temperature & UV irradiation, with and without moisture influence) on PHB films and PHB reinforced with hemp fabric was analyzed by Michel and Billington (2012).The hemp composite exhibited mass loss, swelling, decreased strength and Young's modulus in both processes.Moisture generally enhanced the composites and polymers' degradation, fading and cracking.For the composites, cracking was identified as the most critical mechanism for reducing physical and mechanical properties (Michel and Billington 2012).In our experiments, the interface between fiber and matrix also seems to be the weak point in the composite material.Due to swelling and detachment of the interfaces, degradation and disintegration of the structures can occur particularly effectively there.
In addition to the effects mentioned, the type of natural fiber used must also be considered.Gómez & Michel (2013) generally attribute a slower degradation to natural fibers, but in their study, fibers such as coconut were used, which naturally degrade very slowly due to their high lignin content (Müssig et al. 2010).If the range of chemical composition of natural fibers from high lignin to negligible lignin content is utilized, the disintegration rate of natural fiber composites with PHB should vary significantly.Based on our results, we can show that the choice of natural fiber can significantly increase the disintegration rates.The results show that the lignin content seems to play a major role.All natural fibers with increased lignin content (wood and hemp) showed a reduced disintegration rate compared to fibers with negligible lignin content (cotton).The most significant increase in the rate of disintegration was achieved with animal hair (wool).

Conclusion
Introducing natural fibers is auspicious for improving the disintegration rates of the compostable thermoplastics PHBH.Cotton shows a sizable cost-saving potential due to the high mechanical properties achieved in combination with the low cost.A strong influence of fiber type on the disintegration rate of the composites was observed.The effect was decisive in the samples containing wool fibers, probably due to the nurturing substrate that wool gives to keratinolytic bacteria and a large number of cracks and increased surface area due to exposed fibers on the surface of a relatively weak fiber-matrix interaction.It is also interesting to note that the impact strength was improved relative to the neat matrix.A cost-saving effect could be implemented if a cheaper wool type is used in composite applications without sacrificing mechanical properties, improving toughness and allowing faster biodegradation.It has been shown that the degradability of compostable polymer materials such as polyhydroxybutyrate (PHB) can be controlled by using natural fibers.Cotton is an attractive candidate from different points of view.In addition to good degradability properties, it also showed excellent mechanical properties, especially impact strength.Wool, too, showed exciting potential in the studies by improving the disintegration rate, and a possible fertilizing effect could result in unique advantages in an agricultural environment.Due to the high wool prices for the textile industry, the use of previously unused wool from waste and residual streams is favored, but the influence of these wool types on the mechanical properties and the possibilities for improving fiber-matrix interaction should be investigated.

Highlights
• The disintegration rates of the thermoplastic PHBH can be controlled by using different natural fibers.
• Cotton is particularly interesting for reinforcing PHBH, has good degradability, and excellent mechanical properties.
• Wool-PHBH composites show exciting potential, a significantly improved disintegration rate combined with a possible fertilising effect.• A material index is proposed for the cost-saving potential taking into account costs and mechanical properties.

Figure 1 .
Figure 1.Mechanical properties of wool-thermoplastic composites, wherein tensile strength is plotted against Young's modulus.The dotted ovals encapsulate data points achieved with the same matrix.(IFR = intumescent flame retardant) (Data based on a literature review (Röhl and Müssig 2022).
) φ : Fibre mass fraction; C f : Cost per kg of the fibers; C m : Cost per kg of the matrix; C c : Cost per kg for compounding; ρ : Composite density

Figure 2 .
Figure 2. Stress-strain curves of the PHBH specimens investigated in the tensile tests.

Figure 3 .
Figure 3. Visual inspection of the excavated pieces after a given period.The first picture is brighter due to a different environment.Picture after 1 week of burying should be used as a reference.

Figure 4 .
Figure 4. Relative mass loss of the samples buried in the soil.Coloured areas specify the range of the respective standard deviation; 3 specimens were excavated for each date and material.

Table 1 .
Compositions of the created compounds as well as fiber sources and prices.