Valorisation of chicken feather bres in developing poly (vinyl chloride) biocomposites

The aim of this paper is the valorisation of bres obtained from chicken feathers in developing new poly (vinyl chloride) (PVC) biocomposites. Feather bres were rst characterized by thermogravimetric analysis and scanning electron microscopy. Then, semi-rigid and exible PVC biocomposites were prepared by varying the bre (5, 10 and 15 wt %) and plasticizer (15, 30, 45, 60 wt %) contents. The density, mechanical, thermal and morphological properties of the composites were investigated. The water absorption was determined at two temperatures (23 and 60°C). The results showed the inuence of bre and plasticizer contents on the considered properties.


Introduction
Undoubtedly, polymer composites are an important group of materials with a widespread use. The rising worldwide concept of sustainable development created a new scope for the preparation of new composite materials based on renewable resources.
Within this context, to develop bio-based products or biocomposites, one potential approach is to make composites from natural bres. Recently, a great interest to use natural bres as reinforcement is gaining attention. Wood bres are the most widely used short bres, but they can also be obtained from agro bases from different parts of the plant such as bast (jute, abaca, ax, hemp, kenaf), leaf (pineapple, sisal, screw pine), seed or fruit bre (coir, cotton, oil palm), grasses and reeds (bamboo, sugarcane), etc. [1].
In addition to their environmental friendliness, the other advantages of natural bres include a good stiffness and strength, with at the same time a low density compared with glass bre. The speci c Young's modulus of natural-bre-reinforced composites is comparable with that of glass-bre composites. Natural-bre-reinforced composites have good lightweight construction potential and show positive break behaviour, i.e. they break without rough edges and the components do not splinter [2].
In the eld of using natural bres in polymer composites, chicken feather bres have attracted much attention. Chicken feathers biomass, an available waste of the poultry industry, can be a renewable resource of natural bres. This will allow developing successful applications for poultry feather in composite making. Indeed, inappropriate disposal of these wastes causes environmental pollution.
Chicken feathers are approximately 91% protein (keratin), 1% lipids and 8% water [3]. Keratins are highly specialized brous proteins. They are resistant to physical and chemical environmental factors. They are insoluble in water, weak acids and alkalis, organic solvents, and are insensitive to the attack of common proteolytic enzymes such as trypsin or pepsin [4].
Feather bre is the product from clean sanitized feathers after removal of the quill. It has a crystalline melting point of ca. 240°C [5]. The semi-crystalline and cross-linked structure in keratin feather bre results in a polymer with a relatively high elastic modulus of approximately 3.4-5 GPa [6]. Tesfaye et al. [7] measured the diameter of chicken feathers fractions. The mean diameters of the chicken feather barbules and barb were respectively 4.93 µm and 46.65 µm.
In this paper, chicken feather bres (CFF) were compounded with poly (vinyl chloride) (PVC) resin to prepare bio-based composites. Chicken feathers are a renewable biomass that would be interesting to valorise. The recycling of feather waste allows the preparation of new materials within a perspective of eco-design or sustainable development, and the processing of poultry waste. It will also prevent wasting a useful material and reduce the use of fossil thermoplastics.
PVC is an important thermoplastic material because of its versatility and low cost [18]. It is usually reinforced with inorganic materials such as glass bres, calcium carbonate, and talc. However, these types of llers are characterized as high density materials which might increase the overall density of the composite materials [19].
The aim of this work is to make a blend of conventional and renewable resources, leading to a product that is partially renewable, and then to study the properties of the obtained materials.

Materials
The chicken feathers were cleaned according to a process patented by the USDA [20].They were grounded for 3min by using a grinder KIKA-WERKE type M20 to separate the bres from the quill. A second grinding step was performed for 5min on Waring commercial grinder to obtain bres with lengths inferior to 4mm.
Commercial PVC (K value = 71.5-72.9) type SE-1300 from Shintech Inc. USA was used as received.
Diisodecyl phthalate (DIDP) from Société Générale des Plasti ants (General Society of Plasticizers) (Tunisia) was used as plasticizer. Calcium-zinc (Ca-Zn) complex and epoxidized soybean oil, respectively, from Betaquimica (Spain) and Henkel (Germany) were used as thermal stabilizers while stearic acid from SO.G.I.S.SPA (Italy) was used as lubricant.
PVC and additives were rst mixed in a two-roll mill at 140°C until gelation. Then, bres were introduced and mixed with the matrix for about 15 min at 5, 10 and 15 wt%. The bre content was extended to 20 and 25 wt% for the formulations with 60 wt% of DIDP. Then, the obtained blends were melt compressed at 170°C under a pressure of 300 kN during 5 min, save for the formulations at 15 wt% of DIDP which were pressed at 160°C.The obtained plates were 2 mm thick.
The formulations were designated by a number corresponding to the plasticizer content, followed by a number indicating the bre content. For example, PVC having 15 wt% of DIDP and 5 wt% of CFF was designated by 15/5.

Density analysis
The density was determined by using the pycnometer method. All the reported results are averages of ve measurements for each blend composition.

Water Absorption analysis
Water uptake in composites was determined by immersing weighed amounts of specimens in distilled water at 23±2°C. After 24 h of immersion, samples were removed from water and wiped with lter paper and weighed again.
Furthermore, immersion of the composites in distilled water at 60°C for 30 min and 24 hours was performed.
Where M 0 is the initial mass sample and M 1 is the mass of the sample after immersion.

Tensile testing
Tensile properties were measured using Zwick/Roell testing machine with a crosshead rate maintained at 100 mm/min. Five measurements were conducted for each blend composition and the average value was reported.

Shore D hardness analysis
The Shore D hardness of the samples was undertaken by using a Harteprufer Bareiss durometer according to ISO 868. The reported values of the hardness are averaged from ve measurements on different parts of each specimen.

Morphology analysis
The morphologies of the fracture surfaces were observed by using a JEOL JSM-6360LV scanning electron microscope. The samples were fractured in liquid nitrogen. The chicken feather bre was cut with scissors. The samples were coated with a thin layer of gold before scanning electron microscopy (SEM) examination.

Thermogravimetric analysis
The thermogravimetric analysis (TG) was performed in argon atmosphere using a TA(TGA Q50 V6.1 Build 181) instrument at a heating rate of 15°C/min, from 23 to 580 °C.

Density Variation
The in uence of bres on the density of the PVC composites is shown in Table 1. All PVC compositions lled with feather bres have lower density compared to their matrix. Density decreased with increasing amounts of bres. This can be accounted for the light weight of the chicken feather bres (0.89g/cm 3 as determined by Barone and Schmidt [9]). No natural or commercially available synthetic bres today have a density as low as that of chicken feathers [7].

Water Absorption
The results of the percentages of water uptake in biocomposites are given in Table 2. The data shows that water absorption in the biocomposites is greater than the hydrophobic PVC matrix. It increases with increasing bre content in the biocomposites. Moreover, the water uptake increased for all the samples when the temperature increased, because temperature accelerates the diffusion of water in the material. This output comes from the following facts: 1) water is absorbed by the bres in the composite, 2) the presence of voids or cracks or even poor matrix/ bre adhesion [21].
Chicken feather bres remain susceptible to water absorption according to the amino acid sequence which reveals that keratin has about 40% hydrophilic chemical groups and 60% hydrophobic chemical groups in its structure [22]. The great water absorption in composites can be the result of their increased porosity, which facilitates the penetration and accumulation of water within the biocomposites. Porosity in natural bre composites has been shown to increase with bre content; it is one of the main factors affecting mechanical performance of natural bre composites [23].
The water absorption characteristic depends on the content of the bre, bre orientation, temperature, area of the exposed surface, permeability of bres, void content and the hydrophilicity of the individual components [24].
Tensile Properties Figures 1-3 illustrate the effect of CFF and DIDP content on the tensile properties of PVC composites. According to Reddy and Yang [25],who determined that chicken feather barbs have strength of 180 MPa, elongation of 7.7% and a modulus of 4.7 GPa, it can be noticed that the strength and modulus of bres are greater than that un lled PVC(from Figures 1 and 3), so the CFF are stronger and stiffer than the matrix. Figure 1 shows that increasing amounts of the bre induces decrease in stress at break of the CFF reinforced biocomposites. Furthermore, the increased plasticizer content has also lowered the stress at break in both un lled PVC and CFF/PVC composites, because of plasticization of the matrix. In general, at a higher plasticizer concentration (more than 15-20%), the materials become softer and tougher, with a lower tensile strength, a lower modulus, and a higher elongation at break and a higher impact strength [26].
On the other hand, when bre content increases, the elongation at break decreases as depicted in Figure  2. This shows that bres make biocomposites more brittle; for, the presence of bres in the matrix reduces the ability of the sample to deform by restricting the mobility of the polymer chains. As a consequence, it is di cult for the segments of the material to easily slip past each other [27]. Figure 3 shows that Young's modulus of each CFF/PVC biocomposite is higher than that of un lled PVC, except for the formulations with 15 wt% DIDP. On the other hand, the Young's modulus decreased signi cantly as the DIDP concentration increased. This indicates that CFF improve stiffness of PVC and the plasticizer reduces the rigidity of the composite. This is a common behaviour for the polymers lled with natural bres. Fillers are said to be much stiffer than the polymer matrix, and as a result, they add stiffness to the nal product [28]. Similar behaviour was reported by several authors with different natural llers in PVC composites [28][29][30][31].The decrease in tensile strength with increasing amount of ller was attributed to the poor dispersion of the ller in the PVC matrix and the increase of interfacial defects or debonding between polymer and ller [27]. Moisture pick-up in the bre was also expected to participate in decreasing the tensile strength [32].
Baba and Özmen [13] reported that Young's modulus of PLA reinforced CFF composites is higher, but their tensile strength and elongation at break are lower than pure PLA. In the case of polyethylene matrix, Barone and Schmidt [9] prepared polyethylene/CFF composites at a rate of 0-50 wt%. These authors reported that the elastic modulus and yield stress increase as feather bre loading increases. The yield strain decreases as bre loading increases. Table 3 shows that the Shore D hardness of composites enhanced upon addition of bres. However, the hardness decreased with increasing plasticizer content for un lled PVC and CFF/PVC biocomposites. It can be concluded that the rigidity of the composites is improved with the reinforcement of CFF.

Shore D hardness Evolution
The SEM micrographs of a chicken feather bre are presented in Figure 4 while the SEM micrographs of fractured surfaces of the 15/5, 15/15, 60/15 and 60/25 composites are shown in Figure 5. Figure 4A and B shows, respectively, barb and barbules. It can be seen that the barbules do not have a smooth surface and have hook-like structures along their surface. The cross-section of a feather bre ( Figure 4C) reveals a hollow structure. This latter makes barbs to be very light in weight [25]. The brillar surface of keratin bre is evidenced by Figure 4D.
Barone and Schmidt [9] stated that the intrinsic surface roughness of the feather bres increases the surface area by a factor of about 2.2 over a perfectly smooth bre. The possibility of strong chemical compatibility and lots of available bre surface area may increase the bre/surface interactions over smooth inorganic bres or cellulose-based bres. Figure 5D shows the presence of holes with same diameter of the bres, which demonstrates that the bres were pulled out from the matrix, probably because of a low adhesion between the bre and the matrix. This is con rmed by Figure 5A which reveals the presence of a gap between the matrix and the bre. The insu cient adhesion may be caused by the hydrophilic groups of feather keratin (40%) which have relatively low compatibility with the hydrophobic PVC matrix. It can be also attributed to the presence of water or ethanol residue used to clean the feathers. In addition to that, Figure 5B and C show the appearance of voids that represent the porosity.
However, there are some bres that are still attached to the matrix, which implies the presence of a favourable interaction between the two phases, probably related to the hydrophobic groups of feather keratin (60%) which are compatible with the polymer matrix.
From the images, it can be suggested that there is an interaction between the bres and the polymer and that the bres are moderately dispersed. Good bre dispersion promotes good interfacial bonding, reducing voids by ensuring that bres are fully surrounded by the matrix [23].
These micrographs could explain the decrease in stress at break of the composites by the porosity and the insu cient adhesion between the bres and PVC matrix. Indeed, the observed gap limits the transfer of stress from the matrix to the bres. It is also noticed that some bres are not oriented parallel to the direction of the applied load, whereas the bre reinforcing effect is most e cient along the bre axis orientation [33].

Thermogravimetric Analysis
The thermal stability of CFF and CFF/PVC biocomposites was investigated by thermogravimetric (TG) and derivative thermogravimetric (DTG) analysis ( Figure 6A and B). The corresponding data are given in Table 4.

Page 8/19
TG and DTG thermograms of chicken feather bres display two steps of weight loss. The rst one occurred up to100°C with a weight loss of about 10%. It characterizes evaporation of water and ethanol used to clean the feathers. The second step occurred from 200°C and characterizes decomposition of CFF. The weight loss was of about 70% with a maximum mass change at 323°C. Brebu and Spiridon [34] have identi ed the compounds resulting from the thermal degradation of keratin waste (sheep wool, human hair and chicken feathers), and found the following: inorganic gases (NH 3 , CO 2 , SCS, SCO, H 2 S and SO 2 ), water, thiols, nitriles, aromatics, pyrroles, pyridines, amides, sulphides, thiazoles and thiophenes.
The TG curves of un lled PVC and biocomposites are close to each other because PVC accounted for a high weight in biocomposites. The DTG curves of un lled PVC and composites exhibit two decomposition peaks. The rst indicates a loss of mass occurring above 200°C and the second above 400°C. Concerning biocomposite materials, the rst peak is attributed to the plasticizer migration and the release of HCl with some benzene traces [35] owing to the thermal degradation of PVC matrix, plus the degradation of keratin. The second peak corresponds to the polyacetylene cracking in PVC matrix [35].
As can be seen from the Figure 6, the thermal decomposition of the biocomposites starts before PVC matrix. The temperature at maximum weight loss in the rst step is higher in biocomposites, due to the degradation of CFF at higher temperature, in contrast to the second step where the temperature is lower. Simultaneously, the speed of maximal weight loss is lower in biocomposites.
Additionally, the smooth TG curves correspond to the homogenous loss mass behaviour which indicates homogenous structure of composites [36].

Conclusions
The work presented in this paper shows the potential of bres obtained from chicken feathers as a reinforcement of a polymer matrix (PVC). As a result, the addition of untreated CFF into PVC reduced the density of the product. The biocomposites exhibited more water absorption with respect to the PVC matrix. The stress and elongation at break decreased with the increase of CFF content. However, there was an increase of Young's modulus and Shore D hardness. With increased bre amounts, the resulting biocomposites became more stiff and rigid while being more lightweight.TG analysis revealed that chicken feather bres are thermally stable at the temperatures of PVC processing. Despite DIDP plasticizer reduced the tensile properties; it has improved the processability of CFF/PVC composites.
Finally, the results showed that chicken feather bres can be used e ciently in the manufacture of biocomposite materials based on natural resources, and thus turning waste into resource.