Properties and optimal manufacturing conditions of chicken feathers/poly(lactic acid) biocomposites

1 Pre-print of: J. Cañavate et al. Journal of Composite Materials DOI: 10.1177/0021998315595534 Properties and optimal manufacturing conditions of chicken feathers/poly(lactic acid) biocomposites J. Cañavate1, J. Aymerich1, N. Garrido1, X. Colom1, J. Macanás1, G. Molins1, M.D. Álvarez1 and F. Carrillo1,2 1 EETDepartment of Chemical Engineering, Universitat Politècnica de Catalunya, Colom 1, Terrassa, 08222, Spain. 2 INTEXTER, Universitat Politècnica de Catalunya, Colom 15, Terrassa, 08222, Spain. Corresponding author: Fernando Carrillo Navarrete, e-mail: [email protected], Tel: +34 937398703; FAX: +34 937398225. ABSTRACT Chicken feathers (CFs) waste from poultry industry was incorporated in poly(lactic acid) (PLA) matrix to obtain an environmental friendly biocomposite taking advantage of the unique properties of CFs, such as low density, biodegradability and good thermal and acoustic properties, and of the biodegradability of the PLA. The effect of manufacturing conditions on the final properties of the composite and on the matrix-fiber compatibility was studied. Optimal manufacturing conditions, in order to obtain the best mechanical results, were found at a temperature of 170-180ºC for a processing time of 5 min and a speed of mixing of 50 rpm. Young’s modulus was not very affected by the CF’s content showing a maximal variation of less than 8%, indicating that is possible to include CFs in a composite maintaining its stiffness. However, tensile strength and elongation decreased up to a 58% and 12%, respectively, when CF content was 25% because of the restraining effect of the fibers. Moreover, dimensional stability was negatively affected improves with the inclusion of CFs. Infrared spectroscopy and Scanning electron microscopy studies show that fiber-matrix interaction exists but it is weak. Keywords: green composites; poly(lactic acid); feather, keratin, biocomposite, tensile properties. 2 1. INTRODUCTION Chicken feathers (CFs) are a by-product of poultry industry that is extensively produced in the world. In particular, about 910.000 ton of CFs are annually produced in Europe which are mainly treated as an organic waste, i.e. by incineration or composting. [1] However, due to environmental concerns, their reuse as a material for alternative applications is gaining attention. In this sense, several studies have been developed in order to find applications that valorise their unique properties and that take advantage of their low cost. [2] In fact, conveniently treated CFs waste has already been reused and applied as a biosorbent of contaminants [3], as sound absorption barrier [4] and as reinforcement in biocomposites. [5] Hence, the CFs can be used to prepare materials with interesting properties, for instance for the design of new formulations of composites, even if there is still little information about the different behaviour of fibers, whole feathers and crushed feathers. Moreover, the effect of the different sanitizing pre-treatments has been scarcely considered. [6] Up to now, several studies have been focused in combining CFs with thermoplastics [7] including polyethylene (PE) [8] and polypropylene (PP) [9] and there are also studies involving thermosets such as epoxy resins. [10] The effective compositions of these materials have often involved about 20% of CFs [11] and, regarding their properties, elastic modulus and yield were found to increase when CFs were incorporated in the pure matrix. [12] Besides, PP-based materials reinforced with whole chicken feathers demonstrated a sound absorption coefficient close to one in the frequency range of 44.5 kHz. [9] Differently to conventional thermoplastics such as PP and PE, PLA is a biodegradable thermoplastic that has been used either pure to obtain biodegradable products or as a matrix in composites. So far, in order to take advantage of this feature and produce ecologically improved materials, some natural fibers (mainly of cellulosic nature) have been considered as a priority. Studies on composites including flax [13], bamboo [14] or kenaf [15] are some examples thereof. Following this trend, the combination of the unique properties such as low density, biodegradability and good thermal and acoustic properties of the feathers keratinaceous residua [6] with the biodegradability of the PLA could constitute an advance in the production of more environmental friendly composites. It is important to mention that in most of the previously studied cases, CFs fibers (CFFs) were separated from the quill and subsequently used to prepare the composite, resulting in a biodegradable composite with improved tensile and thermal properties. [5] Alternatively, other previous studies indicated that the addition of whole crushed chicken feathers to PLA preserved the tensile moduli and enhances the biodegradability of the CFs/PLA composites. [16] From the technical point of view, there are several parameters that may influence the performance of composites. On one hand, the manufacturing conditions have an effect in mechanical properties and in the general behavior of the composites. [17] On the other hand, the matrix-fiber compatibility is also an important factor related to the final macroscopic properties of the composite product. [18] It is important to take into account that, in opposition to cellulosic fibers which are mainly hydrophilic, CFs are made of keratin which is a protein with both hydrophobic and hydrophilic amino acids in a ratio close to 3:2 what might be advantageous in regards of materials compatibility. [19] The work developed in this study aims to determine the optimal manufacturing conditions of temperature, mixing time and mixing speed of the process of preparation of PLA/CFs 3 based composites in order to obtain the best tensile properties. To do so, a range of compositions of PLA and CFs were studied and both mechanical and physical properties were determined at several settings. Spectroscopy and microscopic techniques were used to corroborate the mechanical results and to microscopically understand the performance of these materials. 2. METHODOLOGY 2.1. Materials Poly(lactic acid) was supplied by VELOX under commercial name Biopolymers PLE 5 (GMO-free). It was a transparent grade, with a melt flow index of 2 g/10 min and density of 1.25 kg/m3 according to manufacturer’s data. In order to avoid water in the material, PLA was previously dried at 70ºC for 4 hours in oven and subsequently stored in a desiccator. Since CFs from the slaughterhouse were highly biodegradable, it was crucial to sanitize/clean them before their use as technical material. Therefore, CFs were sanitized by means of an autoclave process with steam at 135 ºC for 20 minutes. After that, CFs fibers were dried in an air oven at 60 ºC for 48 h. Deionized water was used in all procedures. To homogenize particle size, clean whole CFs were chopped with a mill machine (RETSCH SN 100 Germany) at a speed of 1500 rpm until each particle size was smaller than 1 mm. Finally, CFs were air-dried at 105ºC for 4 hours and kept under dry atmosphere (desiccator) just before the compounding of the composite. 2.2. Composite preparation Composite specimens were obtained by mixing the previously ground and dried CFs with PLA matrix as described elsewhere. [16,20] In detail, five different compositions were studied: 5, 10, 15, 20, 25% fiber volume fraction (v/v). Furthermore, neat PLA was used as reference. Components were mixed using a Brabender mixer type W 50 EHT PL (Brabender® GmbH & Co. KG, Germany). PLA matrix was melted first and fibers were added later while mixing. The blend was then consolidated in a hot plates press machine type Collin Model P 200E (Dr. Collin GmbH, Germany) forming square plates, measuring 184 x 184 x 2.2 mm3. Consolidation was carried out at a pressure of 100 kN for 5 min at 180ºC composites. Finally, the square plates were cooled under pressure using cool water for 5 min. 2.3. Mechanical testing Tensile tests were carried out in an Instron 3366 (Instron, UK) universal machine following the specifications of the ASTM-D-638-84 [21]. Prior to the test, already prepared composite square plates were properly shaped according to the ASTM 412 [22] specifications. Speed of the test was set at 1 mm/min and temperature and relative humidity were 23 ± 2 ºC and 50 ± 5%, respectively. From load versus displacement test curves, Young’s modulus, tensile strength at maximum load and elongation at break were calculated using Bluehill version 2 software. Up to Eight replicate specimens per sample were analyzed and both average and standard deviation were calculated. 2.4. Experimental design 4 A design of experiments was carried out in order to study the effect of the three main operational parameters (time of mixing in min, speed of mixing in rpm and temperature in degree Celsius) and their influence on the selected mechanical properties (Young’s modulus, tensile strength and elongation at break). Three different experimental values were defined for each parameter in order to perform an accurate study of the influence of these factors; so, a total of 27 experiments were randomly carried out (see values at Table 1). Note that all these experiments were based on the preparation of CFs-based composites containing 20% v/v of CFs, according with the procedure indicated in section 2.2. This composition was considered as representative of the range of materials that present interesting mechanical properties and also useful for comparison with other composites proposed in other studies. [23] Table 1. Experimental design for the preparation of 20% v/v CFs based composites (*Pure PLA was used in this experiment). Three different experimental values were defined for each factor: 170 ºC, 180ºC and 190 ºC for temperature, 5, 10 and 15 min for mixing time and 50, 75 and 100 rpm for mixing speed. Temperature time rpm 170 5 50 75 100 10 50 75 100 15 50 75 100 180 5 50 75 100 10 50 75 100 15 50 75 100 190 5 50 75 100 10 50 75 100 15 50 75 100 170* 5 50 2.4. Scanning electron microscopy Scanning electron microscopy (SEM) microphotographs of composites samples were taken to qualitatively examine the fracture surface of the broken samples to study the 5 fiber/matrix compatibility. The photographs were taken in a JEOL 5610 (JEOL, USA) scanning electron microscope at the accelerating voltage of 10 kV. Samples were previously coated with a 15 nm layer of gold-palladium in order to increase their conductivity. 2.5. FTIR Compatibility Analysis Fourier Transform Infrared (FTIR) spectra were obtained by means of a Nicolet Avatar spectrometer with CsI optics. Powdered samples from prepared composites were ground and dispersed in a matrix of KBr (9 mg of finely divided composite were mixed with 300 mg KBr), followed by compression at 167 MPa to consolidate the formation of the pellet. FTIR spectra were collected in the range of 4000 – 650 cm-1 with 40 scans and a resolution of 4 cm-1. Spectra data were managed by using the software Omnic. 2.6. Water absorption of composites Water absorption of composites was determined by immersion of the specimens in water at 25 ºC for 24 h (ASTM D570-99). [24] First, rectangular specimens (24 x 12 mm) with 2.2 mm thickness were cut from tensile testing fracture specimens and air-dried at 60 ºC for 24 h, cooled in a desiccator and weighed (wo). Then, the excess of water on the surface of the specimens was removed before weighing (w). Four specimens were tested and average and standard deviation were reported in the results section. The percentage of water absorption (WA in %) was calculated using Equation 1: WA=w-wo ( ) wo ·100 Equation 1 2.7. Density of the composite The experimental density (ρe) of each composite was determined by the pycnometer method using isopropyl alcohol as the test liquid (ISO-1183-1) [25]. Three specimens were tested per sample and average and standard deviation were reported. 3. RESULTS AND DISCUSSION 3.1 Tensile properties Some selected classic mechanical properties (namely Young’s modulus, tensile strength at maximum load and elongation at break) of PLA/CFs composites were determined and compared for the different manufacturing conditions following the aforementioned design of experiments. Figure 1 shows those results regarding Young’s modulus for the analyzed composite specimens ordered according to their processing temperature, mixing time and speed of the mixing system. A first observation is that Young’s modulus of pure PLA (3.251 ± 0.095 GPa) was not affected by the presence of the 20 % of CFs. The average values of the different samples (eight replicates) were quite similar taking into account the associated uncertainity and no significant change was noticed compared to the pure PLA modulus. Accordingly, changing the processing conditions did not provoke any significant differences in the Young’s modulus values and this mechanical property can be considered almost the same regardless the processing conditions with values close to the modulus of pure PLA. Consequently, in order to select the optimal 6 manufacturing conditions regarding this property it was necessary to consider other criteria such as simplicity, promptness and lower energy consumption to select the most favorable conditions that were set as 170 ºC, 5 min and 50 rpm. Figure 1. Young’s Modulus (GPa) values for a composite 80/20 v/v PLA/CFs as a function of operational parameters. In X-axis first row stands for mixing speed (rpm), second row stands for mixing time (min) and third row stands for temperature (ºC). Figure 2. Tensile Strength (MPa) values for a composite 80/20 v/v PLA/CFs as a function of operational parameters. In X-axis first row stands for mixing speed (rpm), second row stands for mixing time (min) and third row stands for temperature (ºC). Besides, data regarding the study of the tensile strength are plotted at Figure 2. In this case a general decrease of the tensile strength compared to the plain PLA (50.3 MPa ± 3.6) was observed for all the processing conditions. This behavior has also been observed in some other composites that either did not show very strong interfacial compatibility or, like in this case, when the addition of the reinforcement leaded to the 7 formation of microvoids that affected the cross section of the sample and acted as tension concentrators. It is worth to mention that the decrease can be as high as 70% in the case of the processing conditions at 190ºC, 15 min, 100 rpm, very probably due to the fact that these conditions might produce degradation in the samples caused by the high temperature that was extended for 15 min and combined with a high rotating speed. According to the results, to obtain a higher tensile strength and avoid degradation is preferable to work at lower temperatures (i.e. 170-180 ºC) with short mixing times (i.e. 5 min) and low mixing speed (i.e. 50 rpm). The results of elongation at break as a function of the processing parameters are shown in Figure 3. Compared to that of pure PLA, elongation at break of samples did not significantly change when the addition of 20% v/v of CFs was carried out at the lowest mixing speed and shortest mixing time. On the contrary, a substantial decrease of this property was noticed when increasing mixing time and speed being particularly significant at 190 ºC and 15 min. In this case, the lack of deformation of the fibers hinders the material to strain reducing the elongation of the composite. From the study of the data, optimal conditions could be considered as temperature 170-180ºC (average values were 1.45,1.44 respectively), time 5 min (average was 1.57 compared to 1.36 at 10 min and 1.09 at 15) and 50 rpm (1.48 average value versus 1.27 at 75 and 100 rpm). Figure 3. Elongation at break (%) values for a composite 80/20 v/v PLA/CFs as a function of operational parameters. In X-axis first row stands for mixing speed (rpm), second row stands for mixing time (min) and third row stands for temperature (ºC). The experience accumulated from the experiments described above allowed us to directly discard the highest temperature (190ºC) because tensile strength and elongation at break showed a dramatic decrease due to the degradation of materials that such a high temperature implied. Similarly, after the experiments, it was evident that 15 min of processing was an excessively long time that produced degradation and a decrease in mechanical properties. Summing up, the most favorable conditions that seemed to provide the best results in terms of mechanical properties was a temperature in the range of 170 - 180 ºC, 5 min of processing time and 50 rpm of speed of mixing. Anyhow, as regard to the tensile 8 properties it is noteworthy that the replacement of PLA matrix with 20% v/v of CFs did not produce significant changes in the elastic modulus comparable to the pure PLA ( 3.25 GPa). Conversely, the tensile strength and elongation at break recorded a significant decrease when adding 20% v/v of CFs. 3.2 Dimensional stability and density Dimensional stability related to water absorption of samples was considered for the different processing conditions and the results together with neat PLA are presented in Figure 4. The main finding is that this parameter did not show any clear relationship either to temperature or mixing time or mixing speed. In the literature, the water absorption of pure PLA has been reported to be up to a 1% [26], which basically agrees with the obtained experimental results (0.42 ± 0.12). However, the presence of CFs, which are more hydrophilic than the matrix, increased the water absorption percentage. Almost all the experimental values were in the 1.5 to 2.5% range while two very different and inconsistent values were found. These values corresponded to samples prepared at 170ºC, 15 min and 75 rpm (lowest value) and at 180 ºC, 10 min and 50 rpm (highest). As they did not follow the general trend, they can be attributed to a higher heterogeneity of samples or to experimental error. Besides the dimensional stability, also the density of the composite materials was experimentally determined (Figure 5). Similarly to what happened to water absorption, the obtained values were scattered even if the range of the values was narrower in this case (between 0.9 and 1.3 g/cm3, when averages were considered). Figure 4. Water absorption values for a composite 80/20 v/v PLA/CFs as a function of operational parameters. In X-axis first row stands for mixing speed (rpm), second row stands for mixing time (min) and third row stands for temperature (ºC). 9 Figure 5. Density (g/cm3) values for a composite 80/20 v/v PLA/CFs as a function of operational parameters. In X-axis first row stands for mixing speed (rpm), second row stands for mixing time (min) and third row stands for temperature (ºC). 3.3 Effect of CFs content in the properties of the composites The influence of the CFs content was evaluated by testing composites obtained at optimum processing conditions (180ºC, 5 min, 50rpm) with 5, 10, 15, 20 and 25% v/v CFs. Firstly, typical mechanical properties are shown at Figure 6. As it can be seen in Figure 6a, and likewise the obtained results for different manufacturing conditions, the Young’s moduli of the samples were not very affected by the content of CFs in the range 5-25%. Precisely, when adding 20 or 25%v/v of CFs the modulus diminution was just 8 or 3%, respectively. These results showed the feasibility to obtain a stiff composite adding a reasonable amount of CFs as high as 25%. In the same Figure 6a, it is possible to remark that the effect of the addition of CFs is also not notorious in elongation at break. A decreasing trend might be suggested, which would be reasonable because of the restraining effect of the fibers, but the maximum difference compared to plain PLA is only 12%. By statistical analysis (one-way analysis of variance with  = 0.05), no significant differences were found among the different samples for the elongation whereas when regarding the Young’s modulus, the only sample that differed from the rest was the one including 5% v/v of CFs. Therefore, the elastic modulus of composites including 10-25% v/v of CFs is not statistically dissimilar to that of neat PLA. Conversely; tensile strength, as in similar reported cases [16], decreased with the CFs content as seen in Figure 6b. The maximum decrease was as high as 58% when CFs content was 25% v/v. This behaviour could be explained by the deficient interfacial adhesion and for the presence of microvoids that act as tension concentrators when submitted to tensile strength, as stated before. In this case, a reinforcing effect due to the presence of CFs should be discarded in view of the obtained results. 16 [3] Garcia D, López M and Carrillo F. Chicken feather fibres waste as a low-cost biosorbent of acid Blue 80 dye, Desalin Water Treat 2014:1-4. doi:10.1080/19443994.2014.986531 (in press). [4] Huda S and Yang Y. Composites from ground chicken quill and polypropylene, Compos Sci Technol 2008; 68: 790-798 [5] Cheng S, Lau K, Liu T, Zhao Y, Lam PM and Yin Y. Mechanical and thermal properties of chicken feather fiber/PLA green composites, Compo Part B-Eng 2009; 40: 650-654. [6] Carrillo F, Macanás J, Colom X, Cañavate J, Molins G, Álvarez MD and Garrido N. Use of Chicken feathers waste for the fabrication of composite materials. In: 15th European Conference on Composite Materials, Venice, Italy, 24-28 June 2012. [7] Kiew K, Hamdan S and Rahman MR. Bioresources 2013; 8(2): 1591-1603. [8] Barone JR. Polyethylene/keratin fiber composites with varying polyethylene crystallinity, Compos Part A-Appl S 2005, 36(11): 1518-1524. [9] Reddy N and Yang Y. Light-weight polypropylene composites reinforced with whole chicken feathers. J Appl Polym Sci 2010; 116: 3668–3675. [10] Mishra SC, Nayak NB and Satapathy A. Investigation on Bio-waste Reinforced Epoxy Composites. J Reinf Plast Comp 2010; 29: 3016-3020. [11] Barone JR, Schmidt WF and Liebner CFE. Compounding and molding of polyethylene composites reinforced with keratin feather fiber. Compos Sci Technol 2005; 65: 683-692. [12] Barone JR and Schmidt WF. Polyethylene reinforced with keratin fibers obtained from chicken feathers. Compos Sci Technol 2005; 65: 173-181. [13] Oksman K, Skrifvars M and Selin JF. Natural fibres as reinforcement in polylactic acid (PLA) composites. Compos Sci Technol 2003; 63(9): 1317-1324. [14] Porras A and Maranon A. Development and characterization of a laminate composite material from polylactic acid (PLA) and woven bamboo fabric. Compos Part B-Eng 2012; 43(7): 27822788. [15] Ochi S. Mechanical properties of kenaf fibers and kenaf/PLA composites, Mech Mater 2008; 40(4–5): 446-452. [16] Carrillo F, Rahhali A, Cañavate J and Colom X. Biocomposites using waste whole chicken feathers and thermoplastic matrices. J Reinf Plast Comp 2013; 32(9): 1419-1429. [17] Hull D. An Introduction to Composite Materials, Cambridge University Press, 1981. [18] Bullions TA, Gillespie RA, Price-O’Brien J and Loos AC. The effect of maleic anhydride modified polypropylene on the mechanical properties of feather fiber, kraft pulp, polypropylene composites, J Appl Polym Sci 2004; 92: 3771-3783. [19] Barone JR and Schmidt WF. Polymer Composites Containing Keratin, US Patent, US2005/0148703, July 2005. [20] Colom X, Rahalli A, Cañavate J and Carrillo F. Properties and optimal manufacturing conditions of chicken feathers thermoplastic biocomposites, J of Compos Mater 2015;49(3): 295308. [21] ASTM-D638:2010. Standard Test Method for Tensile Properties of Plastics. [22] ASTM D412:2002. Tensile Strength Properties of Rubber and Elastomers. [23] Bullions TA, Hoffman D, Gillespie RA, Price-O’Brien J and Loos AC. Contributions of feather fibers and various cellulose fibers to the mechanical properties of polypropylene matrix composites. Compos Sci Technol 2006; 66(1): 102–114. [24] ASTM D570-99. Standard Test Method for Water Absorption of Plastics. [25] ISO 1183-1:2004. Plastics. Methods for determining the density of non-cellular plastics -- Part 1: Immersion method, liquid pycnometer method and titration method [26] Finkenstadt VL, Liu CK, Evangelista R, Liu LS, Cermak SC, Hojilla-Evangelista M and Willett JL. Poly(lactic acid) green composites using oilseed coproducts as fillers. Ind Crop Prod 2007; 26: 36-43. [27] CES EduPack software, Granta Design Limited, Cambridge, UK, 2012. [28] Pretsch E. Tablas para la elucidación estructural de compuestos orgánicos por métodos espectroscópicos, Alhambra S.A., Madrid, (1988). 17 [29] Cañavate J, Pagès P, Saurina J, Colom X and Carrasco F. Determination of small interactions in polymer composites by means of FTIR and DSC. Polym Bull 2000; 44: 293–300. [30] Painter PC, Coleman MM and Koenig JK. The Theory of Vibrational Spectroscopy and its Application to Polymeric Materials. John Wiley & Sons Ltd., New York, 1982.