Novel insulation panels development from multilayered coir short and long fiber reinforced phenol formaldehyde polymeric biocomposites

This study investigated about the developments of insulation panels from multilayered coir long and short fiber reinforced phenol formaldehyde polymeric (PF) resin. The lengths of coir long fibers (CLF) were within 3 mm, whereas the short fibers (CSF) ranged from 0.1 mm to 1.25 mm. Four composite panels of 360, 680, 800, and 1000 kg/m3 densities were developed by employing hot pressing technology. The thermal conductivity, microstructural, mechanical, and physical properties of the composite panels were investigated. Perceived thermal conductivity values ranged within 0.046280 (0.000494) to 0.062400 (0.001146) Wm‒1 k‒1of the composites demonstrating superior insulation properties. Moreover, the current study also found that mechanical and thermal properties showed improvement with the increase of density. Low-density fiberboards had the lowest performances compared to high-density composite panels, with the exception of the 1000 kg/m3 density, in which fiber agglomeration occurred. Furthermore, all the developed composite panels display superior potentiality for use as effective insulation materials. The FTIR (Fourier transform infrared spectroscopy) analysis also shows an efficient bonding between the cellulosic coir materials and PF resin. The overall characteristics of the composite panels, especially medium fiberboard, show prominent potential for industrial production units by fulfilling the consumer requirements.


Introduction
The increasing awareness of worldwide environmental pollution has increased the importance of bio-based polymeric composite panels. The use of naturally derived fiber materials as potential reinforcements could facilitate the composite sector with enhanced sustainability [1][2][3][4][5][6][7][8][9]. Among other natural fibers, coir is a crucial renewable fiber material. Coir materials are derived from coconut husks and are produced extensively in Southeast Asian countries. Much research on long coir fiber reinforced composites has been conducted. The bulk of the research focuses on coir reinforced with different thermoplastic (like polypropylene (PP), polylactic acid (PLA), PHBV (3-hydroxybutyrate-co-3-hydroxyvalerate), polybutylene succinate (PBS)) [10][11][12][13][14], thermosetting (like epoxy) [15,16], and cementitious reagents like OPC (Ordinary Portland cement) [17][18][19][20]. Sometimes, nanoparticles (NPs) are also utilized to develop the hybrid nanocomposites in order to improve the performances [21][22][23]. Additionally, NPs are not yet widely implemented for industrial production. Therefore, it is tried to innovate the feasibility of multilayered coir material reinforced PF composites having improved insulation properties. However, research on coir chips, a byproduct generated during coir fiber processing, is not extensive, particularly not on chips with improved mechanical properties. In our previous study, we reported on the development of coir chips (without extracting into fiber) along with long coir materials reinforced with MUF (melamine-urea-formaldehyde) [24]. However, the internal bonding strengths were less than 0.5 MPa, which, depending on density, is the standard for fiber composite panels. The reason for this may be the lowered compatibility between the coconut chips and MUF polymers although the coir materials were pretreated before the fabrications. However, if the chips are extracted into fiber forms, they could provide better reinforcement effects in the composite system, which may increase the fiber to polymer adhesions. Furthermore, the changes in polymeric materials could also provide better reinforcement effects. Coir fibers (short fiber extracted from chips and long fibers directly from coconut husks) reinforced with another potential thermosetting polymers such as PF have not been investigated yet. To our knowledge, this fabrication protocol has not been studied yet. The process involves employing hot-press technology, which provided excellent reinforcement effects with improved internal bonding strengths of higher than 0.5 MPa in the case of medium density panels.
As Table 1 shows, Coir fiber is a prominent lignocellulosic material possessing significant lignin (48.3 ± 1.9%), cellulose (43.4 ± 1.2%), and some other polymers like hemicellulose and ash. Chemically, coir fiber contains a higher lignin proportion than cellulose when compared to other natural fibers [25]. Previously, coconut husks were used extensively for culinary purposes, following the collection of liquid endosperm and copra [26]. The disposal of coconut husks and entire dried coconuts had become a considerable environmental threat as both the husks and the entire coconuts could remain in the environment for extended periods of time before decomposing. The utilization of coconut husks as fiber materials mitigates such challenges and ecological hazards. However, coconut husks could have extensive potential in biocomposite productions. In this regard, science and industry are investing significant attention into utilizing coir materials in gainful ways [15,[27][28][29][30]. For some time, thermoset polymers have played a considerable role in developing fiber reinforced composite products [31]. Thermoset polymers facilitate the composite materials with better mechanical, chemical, and dimensional stability as well as thermal performances by creating a "cross-link" with the polymers [32]. Among different thermosetting polymers, polyester [33,34], epoxy [35,36], MUF [37], and PF [38] are used largely for natural fiber reinforced composite production. PF resin shows higher stiffness, chemical resistance, and excellent insulation properties [39]. In the circumstances mentioned above, PF resin is employed for developing trilayered biocomposite panels through reinforcement with coir materials for insulation performances.
The developed products from coir fibers retain versatile potentiality for being used as the low as insulation material for the housing and construction sector and demanded green consumer products. Moreover, the automotive industries, airspace companies, and different sporting good manufacturers are using coir fiber reinforced composites too. Furthermore, coir fiber reinforced panels are getting also much attention by the composite community. However, the utilization of coir chips extracted into CSF materials with CLF ( Fig. 1) could be an interesting area of research, one that would facilitate manufacturers with a new dimension of potential composite materials. Moreover, thermo-mechanical performances are also dependent on the density of composite materials, which was a main focus for this research. Manufacturers from a variety of industries utilize biocomposite materials with varying densities in products including but not limited to automobiles, furniture, insulation, etc. With the advancements of science and technology, demands on sustainable insulation products from natural fibers reinforced with polymeric resin is also getting attentions. Moreover, research studies on coir fibers extracted from coconut chipreinforced PF resin multi-layered composite panels have not been investigated yet especially for insulation materials. The current study examines the reinforcement effects of different densities to provide more options that are viable for manufacturers. Such a novel fabrication technology could provide superior thermo-mechanical performances on the developed composite panels which are prerequisite to the panel industry.

Materials
For the purposes of this research study, coconut (cocos nucifera) chips and CLF materials were collected from Pro Horto Ltd., (Szentes, Hungary). Prior to biocomposite fabrication, the coir chips were defibrated employing a defibrating machine (VZ 23,412 model, Dinamo Budapest, Hungary). The CLF were cut into nearly 3 mm lengths. Chemco a. s. Co., Slovakia generously supplied the chemical reagent PF required for the research. The characteristics of PF are as follows: reddish brown in appearance, viscose (240 to 380 mPa.s) and liquid, dry matter content (48), density 1210 ± 20 kg/m 3 , alkaline pH ranged from 10 to 12.

Coir fiber preparation from coconut chips and sieving
Coconut husks were carefully defibrated by adjusting the distance between the grinders and grain to protect against fiber damage during processing. The defibrated coir fibers were sieved with a sieve analyzer (ANALYSETTE 3Pro, Germany) to produce uniform fibers before biocomposite panel fabrication. Our previous studies provide detailed descriptions of similar sieving protocols for different natural fibers [24,42]. For this test, 100 g coir fibers were randomly selected from the defibrated materials. The sieved lengths of CSF were within 0.1-1.25 mm (Fig. 2). The maximum fibers (42.7%) were of 1.25 mm lengths, whereas 26.6% of fibers possessed 0.355 mm, 22.5% of fibers 0.1, and 7.4% of fibers 0.8 mm lengths. The defibrated fibers exhibited diversified fiber lengths. Conversely, CLFs were cut into nearly 3 mm length before the composites fabrication.

Production of biocomposite panels
Before starting biocomposite panel fabrication, CSF and CLF moisture contents were measured. Our previous studies discuss the detailed measurement procedures [24,43]. The investigated CSF moisture contents were 3.08%, while for CLF it was 3.18%. Furthermore, PF moisture content was nearly 34% and mat was around 10% (considered for recipe formulation). The CSF, CLF, and PF were measured proportionately as per recipe ( Table 2). The proportion of CSF was 70.2%, CLF 17.8%, and PF adhesive 10%. However, the actual quantity of materials varied depending on different densities of the composite panels (360, 860, 800, and 1000 kg/m 3 ). C@SL1 is recipe 1, whereas recipe 2 is C@SL2, recipe 3 is C@SL3, and recipe 4, C@SL4. The dimensions (400 x 400 x 8 mm 3 ) of the panels were also kept uniform. Initially, CSF and PF were mixed evenly by using a rotating drum in the laboratory. The adhesive was continuously sprayed with a spray gun until the completion of measured PF. An extra 10% of materials were taken in order to ensure that exact quantity would remain after the mixing of adhesive and fibers. Finally, the mixed materials were measured again (at the middle) for every composite panel, and then placed inside a wooden frame of 400 x 400 mm 2 over Teflon paper in the steel plate. However, CLF materials were positioned in the upper and lower parts of the mat (Fig. 3). When the materials were in place, the mat was pressed using another wooden lid. Both the wooden lids and framed boxes were removed, and another steel plate was placed over the mat after two 8 mm steel rods were positioned at the two sides. Finally, the mat in the steel plates was transferred to a hot press machine. The mat was pressed under 7.1, 4.7, and 3.2 MPa pressure; the temperature was 140 °C. The composite panels were pressed for a duration of 40 s (15 s for every 1mm thickness) in three consecutive stages to ensure uniform release of the pressure from the panels. When the  pressing period was completed, the machine was cooled down to normal temperature (around 25 °C) and the produced panel was removed from the machine. All the panels were produced accordingly and cured for 1 day in ambient laboratory conditions.

Characterizations
CSF and CLF moisture contents were investigated using a Kern ULB 50-3 N moisture analyzer manufactured by KERN AND SOHN GmbH Co., Germany, according to the procedures of standard EN 322:1993. Thermal conductivity was characterized as per standard procedures of MSZ EN ISO 10456. Our previous studies provided detailed discussions of different cellulosic material reinforced composites [24,44]. The mechanical properties, in terms of flexural performances and internal bonding strengths, were tested using Instron testing equipment (4208, USA). The testing standard for flexural properties followed the EN 310 standard, whereas the EN 319 standard was followed for internal bonding strengths. The crosshead movement speed for flexural properties testing was 5.0 mm/min and internal bonding strength was 0.8 mm/min. The test specimens were prepared using a circular saw (DCS570N XJ model, Pennsylvania, USA). The morphological characteristics were studied employing a scanning electron microscope (SEM, S 3400 N, High Technologies Co., Ltd., Hitachi, Japan) at different magnifications and 10.0 kV voltage. The chemical components present in the fabricated composite panels were also investigated in terms of SEM mediated EDX analysis. Moreover, the chemical bonding of the panels also performed by FT/IR-6300 equipment (Jasco, Japan) within the wavenumber 4000 to 400 cm -1 .

Results and discussion
Mechanical properties of the developed biocomposite panels Figure 4a illustrates the typical load versus the displacement curves of the developed composite panels from CSF and CLF reinforced with PF resin are illustrated in reference to internal bonding strength properties. All the composite samples showed an elastic behavior until they reached to their maximum peak, which was followed by post-peak softening characteristics as well, especially for the coir fiber inclusion in the composite system [45]. However, the explicit effects of density on the composite materials are clearly visible. After a certain density level, a difference evolved, especially in composite panel 4. This may be because the fibers agglomerated at higher density. Therefore, a sudden, declined peak is observed after composite panel 1, 2, and 3. The highest peak for composite panel 1 was at nearly 692 N.

Morphological properties of the developed biocomposite panels
The surface morphology of the fractured surfaces of CSF and CLF material reinforced PF biocomposite panels are shown in Fig. 5. The investigation on fractured surfaces of the biocomposite panels determine the reinforcement effects on the composite system in terms of fiber pull out. The fracture portions of the matrix clearly exhibit the presence of homogeneous distributions of coir materials in the composite materials.
The tearing of fibers demonstrates a strong fiber to matrix bonding, which is also reflected in their improved mechanical performances. After the reinforcement, surfaces of coir fibers become rougher, resulting in better fiber-to-matrix interactions [48]. Moreover, the fiber to polymer gap did not appear significantly, likely because of the stronger interactions between the fiber and the polymer leading to less void generation during composite formation [49]. Stronger chemical bonding is also reflected in the FTIR and EDX investigations. The overall discussions signify strong and successful reinforcement effects in the composite systems.

EDX characterization of developed biocomposite panels
The elemental quantity of control coir fiber and associated PF bonded polymeric composites were examined further. C and O are the main chemical components of natural fibers, which is also shown in Fig. 6. However, when the coir fibers were reinforced with PF resins, another chemical element, N, is significantly detected in the composite system, in

Thermal conductivity of the developed biocomposite panels
The thermal conductivity of all the composite panels (around 400 x 400 x 8 mm 3 ) were investigated before the samples were cut for other tests. It is well-known that the lower values of thermal conductivity demonstrate higher insulation properties [50]. Thermal conductivity of phenolic resin is 0.29 to 0.32 Wm -1 k -1 in the case of 32 to 64 kg/m 3 density, and it is 0.35 to 0.40 Wm -1 k -1 in the case of 112 to 160 kg/m 3 density [51]. Conversely, thermal conductivity of the coir fibers were 0.058 Wm -1 k -1 for 30 to 115 kg/m 3 fiber density [52]. However, thermal conductivity observed in this current research for the C@SL4 sample containing the lowest thermal conductivity value by 0.046280 (0.000494) Wm -1 k -1 indicating the highest insulation property of the material. Conversely, C@SL1 samples provided the highest value of thermal conductivity by 0.062400 (0.001146) Wm -1 k -1 , demonstrating comparatively poor insulation properties compared to other panels. Moreover, composite panels 2 and 3 also provided moderate insulation properties while their thermal conductivity values are 0.062000 (0.000667) and 0.046280 (0.000494) Wm -1 k -1 , respectively. However, all the composite panels provided superior insulation properties for coir fiber reinforced composites  ( Fig. 7). Furthermore, the density of composite panels varies the distinct thermal conductivity of the panels due to differences in associated densities. In a recent research for coir fiber reinforced panel board, the reported values varied within 0.0547 ± 0.0004 to 0.1205 ± 0.0006 Wm -1 k -1 depending on different fiber lengths [53]. In another study for Calotropis procera fiber reinforced composites, the achieved thermal conductivity was within 0.137 to 0.146 Wm -1 k -1 [54]. The overall discussion demonstrates that the perceived thermal conductivity for this current research provided superior insulation properties.

FTIR analysis of developed biocomposite panels
Moreover, the chemical bonding of the molecules in CSF, CLF, and associated composite panels reinforced with PF resins were also investigated in this current study. The FTIR study is an excellent tool for investigating the formation of covalent bonding, the presence of metallic components, and the detection of functional groups in the materials. The absorption bands appearing at 3368 cm -1 indicate the presence of -OH bond [55] in the polymeric structure of both coir and other composited materials, demonstrating a strong presence of cellulosic coir materials in the matrix system (Fig. 8). The absorption bands at 1610 cm -1 is related to aromatic C═C bonding [56]. However, the peaks after the reinforcements with PF resin show an upward band compared to the control coir materials. Furthermore, the symmetric and asymmetric deformations of alcoholic C-H bond is ascribed by the peaks around 1407 cm -1 [57]. The peak at 1407 cm -1 is most prominent for the materials after PF resin reinforcement due to the assignment of C═C benzene ring of phenolic resin [58]. The peaks around 871 cm -1 are related with CH assignment for isolated H, which is also a characteristic of phenolic resin [58]. Interestingly, the presence of this peak is noticed only for composites materials but absent from control coir. Overall, the FTIR spectrum (Table 4) also demonstrates the strong reinforcement effects of coir materials with the PF resin.  Figure 9 illustrates the moisture content and thickness swelling of the composite panels after 2 h and 24 h of water immersion and drying for the same periods. Both moisture content and thickness swelling showed decreasing trends with the increase of density in the composite panels from 360 to 1000 kg/m 3 . However, other studies also provided a similar explanation for coir fiber/epoxy composites [60]. Natural lignocellulosic fibers contain -OH, -COOH, CO, and so on [61,62] in their polymeric structure, which are hydrophilic in nature. Therefore, natural fiber reinforced composite materials also absorb certain amounts of moisture from the atmosphere; hence, the moisture content of the composite materials should be investigated. Moreover, it is also important to check the thickness swelling to ensure the dimensional stability of biocomposites  Figure 10 presents the thermal stability of multilayered coir fiber reinforced composites. The thermal stability of fabricated hybrid composites from coir materials and PF resin is  also investigated in terms of TGA, which is responsible for certain factors like moisture absorption rate, heat contraction, and thermal expansion [66]. The degradation behavior of fiber and polymeric matrix in terms of weight loss against certain temperature can easily be quantified using TGA. Some extent of weight loss in the materials due to moisture evaporation was observed at the beginning of heat exposure. Initially, a weight loss (around 3 to 5%) is observed around 100 °C in the case of control coir materials and other composited products [19]. However, noteworthy weight loss is observed within 220 to 300 °C (Table 5), likely due to the lower proportions of hemicellulose content in coir materials [67]. Moreover, the degradations around 300 to 400 °C are related with the cellulosic materials [55]. However, after reinforcing the coir fibers with PF resin, a better resistance against heat exposure is observed for all the cases until 356 °C. After that, the composite materials start to degrade more when compared to the control coir materials. The control coir materials had a higher char yield than the composite products. Char yield also shows higher values in higher density panels. This demonstrates that higher density composite materials provide less degradation against temperature.

Conclusions
The current study developed and reported upon CSF and CLF material reinforced with PF resin composite panel of variable densities with superior insulation properties. Coir chips were defibrated and then sieved before the fabrication into composites. Subsequently, long fibers were cut into 3 mm lengths before going into composite production. The study produced tri-layered composite panels where the fibers defibrated from chips were used in the core layer and long fibers in the upper and bottom layers in different proportions. The panels were produced by varying the densities from low to medium to high to investigate the reinforcement effects with PF resin in order to attain significant thermal conductivities. The panels demonstrated improved performances with the increase in densities just except the case of high density at 1000 kg/m 3 . However, all the medium density composite panels displayed internal bonding strengths higher than 0.6 MPa, which meets the requirements of fiberboards in terms of mechanical point of view. The insulation properties of composite panels provided superior performances ranging within 0.0624 (0.001146) to 0.04628 (0.000494) W/(m.K), with a 25.8% increase from the lowest to highest densities. The morphological, EDX, and FTIR studies also demonstrate successful reinforcements between the coir materials and PF resin. This current study would be a benchmark for better insulation properties to the composite panel community.
Funding Open access funding provided by University of Sopron.

Data availability
The datasets needed for reproducing these findings/ used during current research are available from corresponding authors upon reasonable request.

Conflicts of interest
The authors declare that they have no conflicts of interest for the submitted work.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.