Investigations on Dielectric Constant of Coir Powder-Reinforced PVC Composites

ABSTRACT In recent times, there is an intensive growth toward investigation and creation of the green fiber composites due to its abundance and cost-effective, renewable, and environmentally-safe features. Natural fiber-based composites are rapidly replacing synthetic fiber-based composites in electrical engineering applications but being limited due to poor electrical insulation properties. Hence, an attempt is made in this study to improve the electrical insulation properties of coir fiber (powder form)/polyvinylchloride composite by optimizing the variables viz. fiber content (wt.%), particle size (in μm), and chemical treatments using Box–Behnken design. Thus, the results are optimized using the analysis of variance to achieve the low dielectric constant (2.256) for the combination of chemical treatments (triethoxy(ethyl)silane), fiber content (2 wt.%), and particle size (179.5 μm), which are suitable for the electrical insulation products. Additionally, a conformity test is executed and error was found to be 2.54%. Graphical abstract


Introduction
India is the largest raw material supplier of natural fibers as a wide range of it is available through agricultural land and forests (Saravana Bavan and Mohan Kumar 2010).Natural fiber composites, also known as green composites (Mantia and Morreale 2011;Mochane et al. 2021), are widely used in a variety of applications like geo textiles, electric packaging materials (Pan et al. 2017), molded products, sorbents, filters (Zahidul Islam et al. 2022), construction materials, furniture, military (Faruk et al. 2012, Kumar et al. 2019), automotive parts (Sathishkumar et al. 2013a), aerospace industry (Sun et al. 2023, Ahmad, Choi, andPark 2014), structural beams, panels (Keya et al. 2019, Shekar Patil andKalagi 2015), cable, wire insulation, and switch boards (Jayamani et al. 2014, Kuram andKuram 2022).Natural fiber is biodegradable, less health hazardous, and process-friendly, has low CO 2 footprint, is recyclable (Ilyas et al. 2022;Kerni et al. 2020;Lotfi et al. 2021), and has low cost, low density, and high toughness (Kannan and Thangaraju 2022;Sathishkumar, Naveen, and Satheeshkumar 2014).Consequently, it is a preferable alternative to synthetic fibers (Maithil, Chandravanshi, and Chandravanshi 2023).Among natural fibers, coir fiber exhibits various properties such as high availability, high lignin content (Elanchezhian et al. 2018;Sathishkumar et al. 2013ba), low density, high elongation at break, and low elastic modulus (George Adeniyi et al. 2019;Kumar Saw, Sarkhel, and Choudhury 2012;Bongarde and Shinde 2014;Verma et al. 2013).The utilization of natural fiber in polymer composite is well known owing to the electrical insulating properties.An exceptional cable insulator possesses good electrical insulation and physical and mechanical properties.An insulator's performance is improved by lowering its dielectric constant (DC; Nayak et al. 2017).
All the natural fibers are hydrophilic in nature (Dugvekar and Dixit 2022).In order to enhance the electrical insulation properties of composite materials, the interfacial adhesion strength between matrix and natural fiber has to be improved using chemical treatment methods (Nassif 2010;Nigrawa and Chand 2012;Siddika et al. 2015).Therefore, chemical treatments using sodium hydroxide (NaOH; Karthikeyan and Kalpana 2022;Sathishkumar et al. 2013) and silane (Siakeng et al. 2018) and also acetylation (Zaman and Khan 2021) are carried out with coir fiber to remove fiber surface impurities (Sudhakara et al. 2013).Chemical treatments also result in reduced water absorption capacity (Sai Priya, Raju, and Naveen 2014).
The effect of dielectric behavior in the natural fiber-reinforced polymer composites depends upon volume percentages of fiber and resin (Bongarde and Shinde 2014), characteristics of their constituent materials (Verma et al. 2013), polarizability of the material (Khouaja, Koubaa, and B 2021), cellulose in fabric (Mustata and Mustata 2014), and filler content (Nigrawa and Chand 2012).Thermoplastic polymer matrix is reinforced with natural fibers because of its biodegradability, recyclability (Huda, Huda, and Widiastuti 2021), specific strength, corrosion resistance, cost-efficiency, design versatility (Awais et al. 2020), and good electrical insulation (Nayak et al. 2017).Epoxy resin, natural rubber, and polypropylene reinforced with palm sugar (30%) and sisal and coconut fibers (25%), respectively, exhibit low DC (Ngurah Nitya Santhiarsa, Pratikto, and Marsyahyo 2014;Peng et al. 2010;V et al. 2023).The combination of 3 wt.% of boron nitride, 2 wt.% of banana fiber, and a particle size of 3 µm results in a composite material with a DC of 1.14 (Salunke and Gopalan 2022).The DC of nonpolar polymer lies between 1.8 and 2.6 and similarly for plasticized-polyvinylchloride (PVC), it lies between 3 and 5 (Nayak et al. 2017).To improve electrical properties of composites, injection molding process is incorporated to initiate the dispersion of fiber in the resin (Saba et al. 2015, Qaiss, Bouhfid, and Essabir 2015, Sapuan and Yusoff 2015.The response surface method is a tool that implements a statistical approach to optimize natural fibers content in composite fabrication for better properties (Pravitha et al. 2021).Central composite design (Pandiselvam et al. 2022;Penjumras et al. 2015) and Box-Behnken design (BBD;Mohamad Zaki Hassan et al. 2019, Mat Kandar andAkil 2016] are well-known response surface methods.Among response surface methodology, BBD is better because it does not have axial points, has fewer design points, is less expensive to run with the same number of factors, and efficiently estimates the first-and second-order coefficients (Preetha et al. 2023;Sudha et al. 2023).
An experimental plan and further statistical analysis of data with regression model, fitting for each response, were carried out using Expert Design software version 10 (Sathishkumar et al. 2017;Srikanth et al. 2020).Analysis of variance (ANOVA) and regression equation were used in explaining the level of influence for different parameters in composite materials (Ahmad et al. 2023;Aydar et al. 2022;Venkatachalam et al. 2018;Yaghoobi and Fereidoon 2019).
From the literature, it is determined that no investigation has been carried out to analyze the electrical insulation of composite material using PVC and coir fiber in powder form.As a result, a novel study attempt is carried out in this paper to reduce the DC of coir fiber powder/PVC by optimizing the variables like fiber content (wt.%), particle size (in μm) and chemical treatments using BBD.Variables affecting the DC are also investigated.

Materials
Coir fiber (Cocos nucifera) is used as reinforcing material because of its easy availability, low cost, insulating characteristics, and eco-friendly behavior.The coir fiber is supplied from Go Green Products, Chennai, India.
The chemical treatments such as NaOH, triethoxy(ethyl)silane, and potassium hydroxide (KOH) are used to improve the bonding between the polymer and coir fiber and those chemicals are purchased from Sigma-Aldrich Chemicals Private Limited, Bangalore, India.
PVC, a thermoplastic polymer, is used as matrix material because of its multipurpose properties such as low cost, durability, being lightweight, and easy processability.The physical properties of coir and PVC are analyzed and the details are given in Table 1 (Kumar Saw et al. 2012;Jammoukh et al. 2018;Rehab and Ghania 2016).Figure 1 shows the different steps of research work processes.

Design of experiment -Box-Behnken design
The design of experiment/optimization study is carried out based on BBD tool of response surface methodology using Minitab version 16 software.It includes three levels (−1, 0, +1) for the independent variables such as fiber content (wt.%), particle size (in μm), and different types of chemical treatments.The output response is DC.In addition, fiber content [A],  2. Table 3 presents the BBD layout for 15 runs.Twelve face edge points and three replicates of origin can be seen in Figure 2.

Preparation of composites
Coir fiber is cut into small pieces (approximately 1-2 cm) using scissors.Coir fiber is heated for 1 h at 105°C in a hot air oven and then using Pulverizes Mill, the fiber is converted into powder form.Finally, coir fiber powder is sieved into three categories namely 75, 150, and 225 (μm) using a sieving machine as shown in Figure 3.These values refer to average size with a range of ±20 microns.The coir fiber powder contents (2 wt.%, 4 wt.%, and 6 wt.%) in combination with size of particles (75 μm, 150 μm, and 225 μm) are chemically treated using BBD as shown in Figure 4.The following lines explain the chemical treatment procedure.• For triethoxy(ethyl)silane treatment, coir fiber powder is immersed in distilled water with 2 wt.
% silane for 3 h.After treatment, the coir fiber powder is thoroughly washed and then dried using hot oven at 80°C for 48 h (Asim et al. 2016).
• For NaOH treatment, coir powder is soaked in 5 wt.%NaOH solution for a period of 72 h at room temperature.Furthermore, the coir fiber powder is removed from the solution and then washed several times using fresh water.Subsequently, the coir fiber powder is washed with demineralized water.Finally, the coir fiber powder is air-dried for more than 2 days (Krishnaraj Chandrasekaran et al.

Dielectric constant measurement
DC is determined by measuring variable capacitance of standard specimen and test specimen at resonance condition with the aid of DC equipment.Originally, variable capacitor of the test sample is measured.Subsequently, standard capacitor values with and without dielectric material are noted.
Similarly, the same technique is followed for other samples.Then, the average value of DC is calculated using equation (1).DC K is measured at the frequency of 5 MHz in room temperature (Tereshchenko, Buesink, and Leferink 2011) as shown in Figure 5. Rectangular samples of length 115 mm, breadth 95 mm, and thickness 3 mm are used for the study.
K= DC.C1= Conventional variable capacitor's capacity at resonance (maximum deflection).C2= Conventional variable capacitor's capacity at resonance (maximum deflection) including test capacitor with dielectric in it.
C3= The conventional variable capacitor's capacity at resonance (maximum deflection) including test capacitor without dielectric in it.

Dielectric constant by experimentation
The experimental DC values for all 15 samples are tabulated in Table 4. DC of composites ranges from 2.254 to 3.283.The highest DC value is 3.283 for the variables having fiber content (4 wt.%) and particle size (75 μm) and undergoing KOH treatment.The lowest DC value is 2.254 under the conditions which include fiber content (2 wt.%), 225 μm of particle size, and NaOH treatment.Interfacial polarization of a composite depends on coir fiber content (Jayamani et al. 2014).Increase in coir powder above 2 wt.% increases moisture absorption capability leading to increase in DC.Particle size (225 μm) in coir fiber leads to good fiber dispersion with p-PVC.Because of chemical treatment, low DCs are obtained due to less interfacial polarization (Jayamani et al. 2021).The statistical comparative study on DC is carried out using response surface methodology, and responses with respect to the effects of the independent variables (Al-Sharify et al. 2022) are discussed in the upcoming section.

Model fitting and ANOVA for dielectric constant
Using Minitab software, the recorded data are examined.ANOVA is used to analyze the responses for model fitting and to evaluate the significance of the coefficient terms (Pragasam and Mallikarjuna Reddy 2020).
The ANOVA for quadratic model on the DC is presented in Table 5 The P-value of fiber content (wt.%) is 0.05 and the F-value of fiber content is 6.07.Hence it is statically significant.Additionally, P-values of other parameters such as (fiber content (wt.%) * fiber content (wt.%)) is fewer than 0.05.The R 2 value (79.63%) shows that the model is a good fit and the relationship between the dependent and the independent variables is significantly stronger.Hence the model is fit to analyze.
By executing simple regression analysis on the responses, the quadratic model for the DC of the three chosen factors is shown in Equation .(2) Where A is a fiber content (wt.%),B is a particle size (μm), C is a chemical treatment.
From the regression equation ( 2), DC values for all 15 samples are calculated by substituting the corresponding fiber content (wt.%), particle size (μm), and chemical treatments as presented in Table 6.
Experimental and theoretical DCs are compared for all 15 samples.The error (%) is calculated and found to be in range −1.68% to 12.907%, which are tabulated in Table 7. Error is less than 10% except for two samples.Few samples' errors are more than 5%.Samples below average exhibit errors more than 5%.

Response surface 3D plot of interaction
Figure 6 demonstrates the surface 3D plot of interaction among input variables such as fiber content (%), particle size (μm), and different types of chemical treatments on DC behavior.Figure 6a illustrates the surface 3D plot on DC vs. particle size (μm)/fiber content (wt.%).It is inferred that low DC is observed for two combinations: 1.2 wt.% of fiber content/160 μm of particle size and 2.6 wt.% of fiber content/160 μm of particle size.

Main effects for dielectric constant
Figure 7(a) shows that rise in the fiber content from 2 wt.% to 4 wt.%increases DC but also a decrease (2.82 to 2.44) is observed from 4 wt.% to 6 wt.% of fiber content.Figure 7b shows that DC decreases when particle size is increased from 75 μm to 225 μm. Figure 7c reveals that DC is low for triethoxy(ethyl)silane chemically treated FRP (Fiber Reinforced Polymer) composites.When a silanol molecule from the triethoxy(ethyl)silane interrelates with the cell wall of coir powder, silanation occurs.The hydroxyl groups (lignin, hemicellulose, and pectin) existing in the coir powder are reduced, thereby increasing cellulose content.As a result, good interfacial bond is created, which leads to decrease in the DC.Larger amounts of lignin, hemicellulose, and pectin are removed by triethoxy(ethyl)silane compared to other chemical treatments.Hence, the silane treatment leads to lower DC in comparison to other treatments (Asim et al. 2016;Sunthrasakaran et al. 2019).

Pareto charts of the standardized effects on dielectric constant
Figure 8 demonstrates the impact of various factors of the simple regression equation on the DC.The influence of fiber content and particle size shows significant effect on the DC.

Verification and optimization of the model
In Figure 9, response optimization plot for DC is shown, with Y-axis displaying DC and X-axis displaying fiber constant (wt.%), particle size (μm), and different types of chemical treatments.It reveals that the set of triethoxy(ethyl)silane treatment, fiber content of 2 wt.%, and particle size of 179.54 (μm) offers low DC.
To attain good electrical insulation properties, it is essential to recognize the maximum significant of each parameter by satisfying composite desirability (Salunke and Gopalan 2022).Thus, the optimized combination provides low DC as stated in Table 8.Table 9 is a comparative analysis of our optimization process, which displays the error analysis for DC between optimization and experimentation values.Error is 2.54% only, which shows the authenticity of the optimization process.

Conclusion
This paper reports the influence of fiber content (wt.%), particle size (μm), and different types of chemical treatments on DC of coir fiber-reinforced PVC composite.Initially, coir fiber is powdered and then chemically treated (NaOH, KOH, and triethoxy(ethyl)silane) using BBD.Using hydraulic injection molding, chemically treated fiber powders are reinforced with PVC to manufacture the samples.
The manufactured samples are experimentally tested for DC using resonance method.Furthermore, ANOVA and regression analysis are performed using Minitab software to find the effect of fiber content (wt.%), particle size (μm), and different types of chemical treatments on output response.The response surface 3D and main effects plots are obtained to examine the effects of fiber content (wt.%), particle size (μm), and different types of chemical treatments on DC.Errors are calculated among experimental values and regression equation values to find the deviation.
Experimentally, low DC value of 2.19 is attained for the set of fiber content (2 wt.%), particle size (179.545μm), and triethoxy(ethyl)silane.Coir fiber is added to the composite to improve its ability to degrade and be recycled.The composite's green content is increased by its incorporation of coir fiber.It is concluded that with persistent and systematic research on fiber-reinforced thermoplastic polymer composites, there will be good opportunity and better future in electrical applications requiring insulation such as microchips, parts of transformers, terminal, connectors, switches, circuit boards, etc.

Highlights
• BBD approach is used for designing the experiments.
• Injection molding machine is used to fabricate the composites.
• ANOVA is used to identify the relevance of individual process variables.
• Lower DC functions as an insulator for electronic products.
2013).•For KOH treatment, Powdered coir fiber is treated with a solution of KOH (10 wt.%).It is kept in the alkaline solution for 36 h at 30°C.Fiber is completely washed in faucet water and it is neutralized with acetic acid solution (2%).It is again thoroughly washed in water to eradicate the acid sticking.Finally for about 48 h, it is left to dry at room temperature (Srinivasa and Bharath 2013).• Hydraulic injection molding machine is used for fabricating all the samples.The powdered coir fiber and PVC are injected into hopper heated barrel that enhances the heating process together with the shearing action of the screw.The parameters used for manufacturing the natural fiber powder-reinforced polymer matrix composites are injection pressure of 130 MPa, injection temperature at 190°C, injection rate of 20 cm 3 /s, and holding pressure of 190 MPa.The cast of each composite is cured at molding temperature of 35°C, cooling time of 15 s, and holding time 10 s (Fu et al. 2011).Thus, the specimens are fabricated in the shape of rectangular blocks, sized 115 mm × 95 mm × 3 mm.

Figure 7 .
Figure 7. Main effects plot for DC.

Figure 8 .
Figure 8. Pareto chart of the standardized effects.

Table 1 .
Physical properties of coir and PVC.

Table 2 .
Parameters for experimental strategy.

Table 3 .
The arrangement of the BBD.

Table 5 .
ANOVA outcomes for DC.

Table 7 .
Comparison between experiment and regression equation for DC.

Table 8 .
Criterion for optimization.

Table 9 .
Optimized set of variables.