Drought Tolerant Plants’ Fiber and Recycled PET Co-Fibrous Composite as Acoustic Absorbers and Thermal Insulators

ABSTRACT  In this paper, the series of co-fibrous composite materials combining natural fiber derived from drought-tolerant plants (pineapple, hemp, sisal, and agave) and recycled-polyethylene terephthalate (r-PET) were successfully prepared using a mixing-hot-pressing method. The area density and thickness of prepared co-fibrous materials were controlled and porosity was calculated. The co-fibrous materials with higher porosity showed better performance in sound absorption and thermal insulation. Furthermore, all prepared co-fibrous materials have a noise reduction coefficient (NRC) higher than the high-efficiency sound absorber standard at 0.56, making them comparable to commercially available products. Regarding sound absorption performance at a high frequency of 2–5 kHz, the prepared co-fibrous materials exhibited exceptional sound absorbing performance with a sound absorption coefficient (SAC, αH) ranging from 0.94 to 0.97. In terms of thermal insulation performance, both pineapple and hemp co-fibrous materials demonstrated a low thermal conductivity value of 0.029 W/mK, placing them in the range of commercial polyurethane (PU) insulating materials. In conclusion, the pineapple and hemp co-fibrous composite demonstrate potential as alternative eco-friendly commercial sound absorbers and insulators that can subsidizing petroleum-based products.


Introduction
For several decades, the rapid development in urban areas has created favorable economic conditions and increased population density significantly.Due to the effect of overpopulation, a series of noise pollution problems have become undeniable and the main sources of noise pollution up to 70% came from traffic followed by social life and construction (Shen, Li, and Yan 2021).According to a study by World Health Organization (WHO), exposure to excessive noise results in serious health hazards (Berglund et al. 1999).Thus, sound-absorbing materials are employed in buildings, automobiles, and aircrafts to secure human well-being from unnecessary environmental noise.This passive noise cancellation methodology aim to reduce acoustic noise by avoiding multiple reflections from the boundaries of enclosures (Beranek 2007;Sagartzazu, Hervella-Nieto, and Pagalday 2007).The effectiveness of acoustic materials is evaluated by the percentage of sound absorbed and the percentage of sound reduction is referred to the noise reduction coefficient (NRC).In buildings and vehicles, pure metals and their alloys exhibit significant values of sound absorption coefficient at high frequencies (SAC, α H ) of tens of kHz but overall have the low NRC value in the range of civilian and industrial applications at 20-2000 Hz (Berglund, Hassmén, and Job 1996;Xi et al. 2011).Commercial synthetic polymers such as polyurethane, polyethylene, and polypropylene foams, have been used as effective porous sound absorbers in the 20-2000 Hz spectral regime.The NRC at a frequency less than 2000 Hz was observed in all of the cases, but among them, only few cases have met the practical standard for sound-absorbing materials.Furthermore, employing porous sound-absorbing materials also has the potential for improving the energy performance in buildings.In general, these synthetic materials contained a solid matrix with interconnected pores, suitable for various purposes in adsorption systems, thermal energy saving systems, and insulation systems (Rashidi, Esfahani, and Karimi 2018).Due to efforts in sustainable development, there has been an increasing interest in developing plant-derived fibrous materials to subsidize petroleum-based plastics for acoustic insulation.Natural fibers are known for their green characteristics, such as low density, excellent sustainability, biodegradability, abundance and renewability, recyclability as well as being nontoxic (Arenas and Crocker 2010).
A number of reports have focused on developing sound absorbers from plant-based fiber such as jute (Shen, Li, and Yan 2021), sunflower stalks (Mati-Baouche et al. 2016), coconut fibers (Taban et al. 2019), bamboo fibers (Koizumi, Tsujiuchi, and Adachi 2002), pineapple fibers (Do et al. 2021;Suphamitmongkol et al. 2023), hemp fibers (Liao, Zhang, and Tang 2022), agave fibers (Mvubu, Anandjiwala, and Patnaik 2019), and sisal fibers (Dhandapani and Megalingam 2022).In term of acoustic performance, the sound-absorbing composite of porous materials is affected by the porosity, the air back cavity, the composite thickness, flow resistivity, density, and tortuosity (Mvubu, Anandjiwala, and Patnaik 2019;Rakesh et al. 2021).Various models of sound absorption mechanism have been used to predict the sound absorption coefficient based on empirical model and nonempirical models depending on airflow resistivity measurements of porous materials and other nonacoustical parameters such as Delany-Bazley, Miki (Raj, Fatima, and Tandon 2020;Soltani et al. 2020) and etc.
Concerning the thermal properties of natural and co-fibrous composites, the main factors influencing these properties were fiber and matrix types, filler addition, fiber content and orientation, fiber treatment, and manufacturing process.The thermal stability of the natural fiber can be enhanced by removing some hemicelluloses and lignin components (Neto et al. 2021).However, fibers with lower lignin content and also fewer stable constituents (i.e., pectin, waxes, and water-soluble substances) are tolerant to thermal damage.Thus, the thermal resistance of the natural fiber leads to critical issues to use for thermal insulator.
In this study, the natural fiber and recycled PET co-fibrous materials have been developed based on drought-tolerant plants, which are pineapple, hemp, sisal, and agave, using a mixing-hot-pressing method.The average noise reduction coefficient (NRC), and sound absorption coefficient (SAC) at low α L , medium (α M ), and high (α L ) frequency of each plant-based composites were investigated together with thermal conductivity and compared to a commercial co-fibrous sound absorber.The objective of this study was to develop alternative plant-based sound absorbers and thermal insulators with high efficiency.

Materials
All natural fibers (pineapple, hemp, agave, and sisal) were extracted manually using a decorticating machine by a group farmers from various districts in Thailand.Pineapple and agave leaves were extracted for fiber by a group of farmers in Ban Kha district, Ratchaburi, Thailand.Hemp fiber obtained from stems extracted by a group of farmers in Phop Phra district, Tak, Thailand.Sisal fibers were produced from leaves extracted by a group of farmers in Cha-am district, Phetchaburi, Thailand.The extracted fibers were washed in water and sun-dried and stored for the subsequent study.

Natural fiber preparation and fiber characterization
All natural decorticated fibers from different sources as mention above were cut to be staple fiber at a cut-length of 50 mm (Figure 1).Fiber characterization was analyzed as following.Morphology of fiber cross-section was observed by สรเ microscope (Leica, LM750, Germany) at 200X and by a scanning electron microscope (SEM) (FEI Quanta 450, Hillsboro, OR, USA) with accelerating voltage of 2 kV.Fiber diameter distribution was analyzed following ISO 5079-1955(E) by light microscope (Leica, LM750, Japan) at least 50 photos.Physical properties of fiber which were tensile strength, elongation, and Young's modulus of fibers were measured following ISO 1579ISO -1955(E) (E) with Autograph (AGS5kN, Shimadzu, Japan) by the average of 50 samples per each fiber with the crosshead speed of 3 mm/min.

Preparation of natural fiber and recycle PET composite materials
The composite materials were prepared from the mixture of each fiber (pineapple, hemp, agave, and sisal) mixed with 7-denier hollow recycle-PET (r-PET, supported by Angtai Co. Ltd., Thailand) and low melting point PET (L-PET, supported by Thai Nonwoven Co., Ltd., Thailand) in the ratio of 30:60:10, respectively.L-PET has lower melting points, that is between 110°C to 180°C than other r-PET.The mixtures were carded with roller carding machine (Mesdan, 3374, Italy) to obtain the good distribution mixture then the mixture was hot-pressed by stenter machine (MD, China) at 150°C for 6 min (Figure 1).This process was to set in order to use L-PET fiber to melt and bonded r-PET and other natural fibers together.The area density of composite materials was controlled to 2.25 kg/m 2 , with a thickness of 5 cm (ASTM D 2736-95).The porosity of each composite material was calculated from ϕ = 1 -p m / p f , where p m and p f are the densities of composite material and the density of the combination of fiber substance respectively (see Supporting S6).Thermal conductivity of each composite material was analyzed with Heat Flow Meter Instruments (HC-074, EKO instruments, Japan) following the standard method ASTM-518-1.Sound absorption of acoustical composite material was measured following standard method ASTM E-1050 by Acoustic Material Testing Single-Sided (Type 3560 B-T72/X72, Bruel & Kiaer, Denmark).

The NRC measurement
According to the standard method ASTM E-1050, each composite sample was prepared by cutting into a circular shape (diameter 3 cm and 10 cm, 2 samples each).Figure 2 shows impedance tube set up measurement which contains an impedance tube, a sound amplifier, a digital frequency analysis system and a computer.Samples with each 3 and 10 cm diameters were put at one end of the impedance tube with the backing plate.During the test in a two-microphone impedance tube, a sound amplifier was mounted at one end of the impedance tube.Then, the sound amplifier produced sound signals propagate as planar waves in the tube and passed through the sample surface.The reflected wave signals were acquired and compared to the incident sound wave.The normal-incident SAC was measured by the two-microphone impedance tube method over a frequency range of 0 Hz to 6000 Hz.The sound signals were collected and normal-incident SAC was calculated by PLUSE TM software using equation 1 following standard test method for impedance and absorption of acoustical materials ASTM E1050.Where H 12 is the transfer function of the sound signal from microphone-1 to microphone-2, j is the imaginary number ffi ffi ffi ffi ffi ffiffi À 1 p , and k 0 is the frequency dependence wave number of sound.In our study, the displacement between each microphone (s) was 0.037 m and the displacement between microphone-1 to the sample (x 1 ) was 0.37 m.

The thermal transmission measurement
According to standard method ASTM-518-1, HC-074 (EKO Instruments, Tokyo, Japan), a heat flow meter apparatus for testing materials in the conductivity range of 0.005-0.800W/m K, was used for thermal transmission measurement in this study (Figure 3).The device consisted of a cold plate and a hot plate which incorporate heat flow transducers for measuring heat flow.The prepared composite samples were cut into rectangle (20×20 cm 2 ) then the samples were put between two parallel plates of heat flow meter instruments at constant but different temperatures.The device was calibrated against flat materials which have a known thermal conductivity with a wide range of thicknesses and conductivity.The Fourier's law of heat conduction was used to calculate thermal conductivity and thermal resistance.The thermal conductivity (I) was calculated following equation 2.
Where QU and QL are the heat flux transducer at the upper and lower plate, respectively.L is the thickness of the testing specimen and ΔT is the temperature different between surfaces of a testing specimen.

Physical and mechanical properties of natural fibers
Due to the different compositions of each natural fiber (see Supporting S1-S3), morphology analysis (see Supporting S4 and Figure 4), physical and mechanical properties (Table 1) of each natural fiber were required for the comparison.The physical and mechanical properties were corresponding in range of the previous researches (Li, Tabil, and Panigrahi 2007;Wambua, Ivens, and Verpoest 2003).
Figure 4 shows general morphology of each bundle fiber under light microscope with different diameters, shapes, and porous structures.SEM images of all natural fibers show distinctive different each single fibers in bundle fibers with porous structure (Figure 5).Concerning to fiber diameter, they can divide into two groups of fibrous composites, fine and coarse natural fibers.Pineapple and hemp fibers were fine fibers.The porous structure of bundle fibers for these fibers were both in partly individual single cell and among the bundle fibers (Figure 4a, b).Hemp fibers had more irregular, ellipsoidal, or polygonal shape of single fibers with thick cell wall and irregular porous structure in bundle fibers which were corresponding to morphology of hemp fibers reviewed by Jiang's group (Jiang et al. 2019).But agave and sisal fibers (the same plant species) were coarse fibers with individual porous structure of each single fiber and thin wall (Figure 4c, d).Due to the smallest fiber diameter distribution of pineapple fiber (see Supporting S5), its uniform fiber gave the highest tensile strength at 303.5 MPa.For the rest of the natural fibers, the fiber with a thicker diameter and more fiber diameter distribution (hemp, and agave respectively) showed a decline in tensile strength.Unexpectedly, sisal fiber, with the highest diameter at 153.8 µm and high fiber diameter distribution, has comparable tensile strength with pineapple fiber.It is suggesting  that the biggest uniform tubing structure of sisal fiber can also increase tensile strength in exchange for fiber flexibility.These properties were corresponding in range of the previous researches (Li, Tabil, and Panigrahi 2007;Wambua, Ivens, and Verpoest 2003).

Properties of natural fiber and recycled PET co-fibrous composite
To make fiber reinforced composite materials, the high tensile strength and high Young's modulus fiber were more preferable to increase the composite strength and flexibility.Also, the fiber with a fine and diameter distribution exhibited better dispersity of fiber in r-PET fiber, requiring less time for carding and producing a uniform composite.With these conditions, pineapple leaf fiber, with the smallest fiber diameter at 46.5 µm, is the most suitable fiber for the reinforced composite.Morphology of pineapple and r-PET fibers in the composites showed in SEM images (Figure 5).With binding of L-PET fiber between two small fiber diameters of pineapple and r-PET fibers as shown with yellow arrows of junction zone presenting porous structure (Figure 5 a-b), the composite structure resulting in more porous and higher homogeneity of porosity than the two fibers with different fiber diameters of agave or sisal fibers and r-PET fibers.
The fiber reinforced sound absorber composites were prepared using natural fiber in Table 1 mixed with r-PET fiber using L-PET as adhesive in the ratio of 30:60:10.The sample name "Fiber-C" represents the mixture of each fiber and r-PET composite material.As a reference, pure pineapple fiber sample, without PET and adhesive, was also hot pressed (pineapple-P).During the hot-pressing process, the composite materials with an area density of 2.25 ± 0.15 kg/m 2 and thickness of 50 ± 10 mm were obtained.The density of each composite material was calculated by the ratio of area density and thickness (see Supporting S6).In general, the noise reduction coefficient (NRC) was related to the porosity of sound absorber.The calculated porosity of each composite and sound absorption properties was shown in Table 2 (Chanlert, Jintara, and Manoma 2022).

NRC and SAC of co-fibrous composite
It is well known that porosity is the crucial characteristic of sound absorption composites (Witczak et al. 2021).To compare the effectiveness of each sound absorption, the noise reduction coefficient (NRC), the rating system in the commercial sound absorber, is required.The NRC is the arithmetic average of the sound absorption coefficient (SRC) at frequencies 250, 500, 1000, and 2000 Hz octaves (Figure 6).According to the standard, practical sound-absorbing materials require to have NRC value equal to or higher than 0.40, while high-efficiency materials require 0.56 (Liu et al. 2020;Samuel et al. 2021).With this standard, all of the fiber reinforced composite materials have met the requirement of high-efficiency with NRC more than 0.56, and even higher than commercial sound absorber PU form (Table 2).From all prepared composites, pineapple-C has the best NRC value at 0.61 due to the highest porosity at 0.970 followed by hemp-C, agave-C, and sisal-C with lower porosity, respectively.Without co-fiber and adhesive, pineapple fiber (pineapple-P), which was hot-pressed in the same manner, demonstrated underperformance in sound absorbing efficiency with NRC 0.38 despite having a high porosity at 0.970, the same as composite sound absorber pineapple-C (Putra et al. 2018;Thilagavathi et al. 2020).It is because PET fiber has more NRC value than natural fiber but, without reinforcing co-fiber, we have encountered the difficulty of hot-pressing r-PET and L-PET adhesive at the same ratio (90:10).
Thus, a minimum of 35 wt.% of L-PET is required to prepare hot-pressed r-PET material and become out of the scope to reduce the amount of plastic in co-fibrous sound-absorbing materials.Co-responding to the porosity of the prepared composite materials, porous materials also can suppress thermal transfer through solids by reducing the cross-sectional area and   increasing the tortuosity of the heat transfer pathway (Collishaw and Evans 1994;Zhao et al. 2018).
According to Table 2 and Figure 7, a commercial sound absorber (Comm-1) with a composition of 70 wt.%PET and 30 wt.% L-PET has the highest NRC at 0.70 which is high NRC following the recommendation index of ASTM C423-22 standard.So, further investigation is required to compare not only overall NRC but also SAC at low, medium, and high frequency to evaluate the prepared co-fiber composites.The SAC range at low, medium, and high frequency was the average sound absorption coefficient measured at each specific frequency defined by ISO 11,654 (1997) standard: low-frequency SAC (α L ) is at 100, 125, 200, 315 and 400 Hz, medium-frequency SAC (α M ) is at 500, 630, 800, 1000, 1250, and 1600 Hz, and high-frequency SAC (α H ) is at 2000, 2500, 3150, 4000, and 5000 Hz, respectively.
According to Figure 7, the commercial sound absorber (Comm-1) has the highest NRC value (0.70) because the NRC was calculated from the average of SAC at low to medium frequency at 250, 500, 1000, and 2000 Hz.Due to this reason, the commercial sound absorber, Comm-1, with the highest α L (0.3) and α M (0.85) has a slightly better sound-absorbing performance than the prepared sound-absorbing composite materials.However, the soundabsorbing performance of the commercial product Comm-1 dropped significantly in the range of 2000-5000 Hz which is the most sensitive frequency range of the human auditory system (Gelfand 2016).The prepared composite materials (pineapple-C, hemp-C, agave-C, and sisal-C) have shown a superior sound-absorbing performance at high frequency with α H more than 0.94 and up to 0.97.Because of the exceptional α H of selected natural fibers in this study, such as Pineapple-P (α H 0.86), the prepared sound-absorbing materials that have overall NRC improved from the r-PET sound-absorbing property also have α H improved significantly.
The porous co-fiber sound absorbers from selected natural fibers (pineapple, hemp, agave, and sisal) with recycle PET have been developed and have met the requirement of highefficiency sound absorber standard comparable to commercial product.By considering the density and the thermal conductivity, pineapple and hemp co-fibrous sound-absorbing material could replace commercial PU foam practically in exchange for increasing the weight density only around 1.5 times (Li, Cao, and Zhang 2008).The use of recycle PET and eco-friendly natural fibers in this study has the potential to reduce environmental impact by subsidizing petrochemical-based PU and becoming alternative absorber and insulator choices.

Thermal conductivity of co-fibrous composite
Due to the high cellulose content of pineapple fiber and its low lignin and hemicellulose content, along with its fine fiber structure, the thermal conductivity of pineapple-C was found to be the lowest compare to the other fibrous composite (Neto et al. 2021).Thus, the thermal resistance of the pineapple-C leads to critical issues to use for thermal insulator.
The low thermal conductivity of pure natural fiber (pineapple-P) also contributes to the improved thermal insulation properties of the prepared co-fibrous sound-absorbing materials.Among the cofibrous composites, pineapple-C with the highest porosity exhibits the lowest thermal conductivity at 0.0195 W/mK, which is even better than the thermal conductivity of glass-fiber batt at 50 mm (0.033 W/ mK).Following pineapple-C, the thermal conductivity values increase for hemp-C, agave-C, and sisal-C, respectively (Table 2).In the case of pineapple-C, the high porosity leads to a decrease in the energy of the incident sound wave due to thermal friction within the air molecules and pineapple pore interiors.The thermal loss and lowest thermal capacity are associated with the viscous loss within the pineapple pores.These two phenomena significantly contribute to the acoustic properties of the porous materials made from both pineapple and r-PET fibers (Cao et al. 2018).In this study, both pineapple-C and hemp-C exhibit thermal conductivity values comparable to those of commercial foam PU and can be used as insulators (Jelle 2011;Wu, Sung, and Chu 1999).Compared to other research on composite materials from cardboard waste and abandoned natural fibers for the manufacture of local acoustic absorbers, sound insulation panels, and thermal insulation panels (Ouakarrouch et al. 2022), our fibrous composites demonstrate higher sound absorption and lower thermal capacity.The composite panels developed in this study are promising candidates for the development of local materials with sound and thermal insulation properties required for building applications.

Conclusions
Co-fibrous materials with high porosity tend to have a high SAC and low thermal conductivity.The co-fibrous materials of recycled PET fiber and natural fiber (pineapple and hemp) have met the highefficiency standard for sound absorbers and insulators.The sound-absorbing performance of the prepared co-fibrous materials, including pineapple, hemp, agave, and sisal with recycle-PET is slightly lower than that of a commercial PET-fibrous sound absorber.However, the prepared co-fibrous materials exhibit excellent sound-absorbing performance within the range of frequencies most sensitive to humans (2000-5000 Hz), with a sound absorption coefficient of more than 94%.Furthermore, the thermal properties of pineapple and hemp co-fibrous materials show a low thermal conductivity at 0.029 W/mK, which is comparable to commercial polyurethane (PU) and standard glass fiber batt insulation.For future studies, alternative approaches to subsidize petrol-based PET fiber need to be investigated, along with exploring additional variations of adhesive to obtain more sustainable products.

Highlight
• Co-fibrous materials of all drought-tolerant plants' fiber in this project and recycled PET have the potential to be high-performance sound absorbers replacing petroleum-based PU. • Co-fibrous materials prepared from pineapple and hemp fiber have a Noise Reduction Coefficient (NRC) of 0.61 and 0.58.• At high frequencies between 2000-5000 Hz, co-fibrous materials prepared from pineapple and hemp fiber has better sound-absorbing performance than commercial fibrous PET sound absorber with a Sound Absorption Coefficient (SAC) of 0.96 and 0.97 respectively.• Co-fibrous materials prepared from pineapple and hemp fiber have the potential to replace petroleum-based PU thermal insulators with thermal conductivity of 0.029 W/mK.• Co-fibrous materials of all drought-tolerant plants' fiber in this project is as light as PU foam with a density of only around 1.5 times.

Figure 1 .
Figure 1.Preparation of the composites with different natural fibers.

Figure 2 .
Figure 2. a) impedance tube set up measurement b) diagram of NRC experiment.

Figure 4 .
Figure 4. SEM images of a) pineapple b) hemp c) agave d) sisal and e) PET fiber cross-section with accelerating voltage of 13-15 kV at a magnification of 1000 × .

Figure 5 .
Figure 5.Light microscope images of pineapple and r-PET fiber on composite surface a) and SEM images of pineapple and r-PET fiber on composite edge with accelerating voltage of 15 kV at a magnification of 300× b).

Table 1 .
Physical and mechanical properties of natural fibers and PET fiber.

Table 2 .
Composite material sound absorption and insulation properties.