Control of biodegradability in a natural fibre based nanocomposite as a function of impregnated copper nanoparticles

Abstract

The properties of biodegradability or non-biodegradability are highly important for the design and application of sustainable materials. The objective of this work is to introduce a novel type of nanocomposite comprising a natural fibre reinforcing agent impregnated with copper nanoparticles, and to control the biodegradability of this sustainable material using the function of the impregnated copper nanoparticles. At room temperature, copper nanoparticles were synthesized and impregnated into palm oil fibre to improve the strength and durability of the fibre and the material properties of the composite. Fourier transform infrared spectroscopic techniques were used to characterize the prepared composites. The biodegradability (minimum to maximum boundary) of the composite was studied (using the soil burial test) as a function of the quantity of copper nanoparticles, where the tensile strength was fixed at the maximum. The property of biodegradability was also optimized with the help of response surface methodology. The biodegradability of the developed composites ranged from 26.72 to 6.51% when the concentration of impregnated copper nanoparticles was varied from 0 to 2590 μg g⁻¹ respectively. The results indicate that copper nanoparticles can be considered as a potential biocide in composite materials and in this work for controlling the biodegradability of the material by varying the quantity of impregnated copper. The relationship between the responses and variables selected in this study has been justified by the predicted models. Moreover, the model terms have been explained and the prediction has also been performed successfully. Thus, copper nanoparticles have been successfully applied for controlling the biodegradability of the nanocomposite materials. The prepared nanocomposite materials are considered for both indoor and outdoor applications. This study is quite promising for controlling the biodegradability of advanced materials, especially when the degree of biodegradability is so important for their respective applications.

---

1. Introduction

Some significant advancements have been achieved globally in industrialization and the economy during the last three decades. The glorious achievements of these advancements are an increase in life expectancy (up to 20 years) especially in developing countries, reduction of infant mortality rates (up to half), improved per capita income, enhancement of literacy rates and the fact that food consumption as well as its production have superseded the population growth rates, etc.^1,2 However, most of these advancements disregard environmental concerns, leading to the conservation of a sustainable world being ignored for future generations. To handle this complex situation, the adoption of biodegradable and sustainable materials should be welcomed by masters in the field.^1–4 Consequently, the requirement of fuel economy and the need to lower pollutant emissions in manufacturing and transportation are increasing the demand for novel-materials which are high-performance (superior mechanical properties), light weight, can replace metals and are definitely low-cost. Nanocomposites are a novel class of polymeric materials, which can suitably meet the increasing demands and even successfully replace metallic equipment in automobiles and aircraft.^3,5,6 For instance, in the past decade, natural fibre based composites have been utilized as door panels, seat backs, headliners, package trays, dashboards, and interior parts which have attracted much attention from European car companies.⁷ The nanocomposite clay–nylon-6 was the first commercial nanocomposite which was used as the cover of a timing belt and under hood materials by Toyota Motor Company.^8–10

The usage of composites is not only limited to automobile parts but it also extends to current generation military aircraft. These advanced materials have the potential to contribute to aircraft development, since one-third of the structural mass (such as critical and primary structures like wing, fin, control surfaces and radome) of aircraft is made from fibre-reinforced polymer composites.¹¹ Indeed, the design and fabrication of composite based aircraft-structures are expanding day by day. Besides, nanocomposite materials are also notable in biomedical, packaging and water purification applications in addition to other sustainable and clean applications.

The term biodegradability indicates a material’s degradation, absorption and clearance after their functions are completed.¹² Due to their wide variety of biomedical applications (tissue engineering, drug delivery and medical devices) biodegradable polymers have been considered to be a focal research material over the past few decades.^13–15 Among the different biodegradable polymers, polyesters provoke the utmost fascination owing to their degradation through hydrolysis and enzymatic methods.¹⁶ The above points of view indicate the significance of controlled biodegradable nanocomposites especially, for sustainable material design and application. Moreover, controlled biodegradable materials are really important for sustainable development.

Based on their applications, it is highly important to control the biodegradability of these materials as some of them are considered for heavy-duty purposes with long lifetimes and some others are developed for short durations. To the best of our knowledge, a detailed study in which a material’s biodegradability is controlled using the function of impregnated CuNPs has not been reported in the literature. So, there is a demand for the ability to control a material’s biodegradability using a novel approach. Consequently, in this study a novel approach was followed for the first time to develop nanomaterials (using low-cost unsaturated polyester resin and OPF, both of which are available in Malaysia) and to control their biodegradability as a function of the amount of impregnated CuNPs. The nanomaterials have been prepared from the CuNP impregnated OPF reinforced unsaturated polyester resin. Here, CuNPs were used in the composite for controlling the biodegradability of the composite as well as improving the strength of the palm oil fibres. The characterization and material properties of the prepared nanocomposites were also addressed by using FTIR and universal tensile testing. The properties of the prepared nanocomposites that were focused on are mechanical strength and biodegradability, which are highly dependent on the quantity of impregnated CuNPs, and the latter of which is controllable by varying the quantity of impregnated CuNPs. The important properties of the developed nanocomposites have been analyzed with the help of RSM and the process for the development of strong and durable nanocomposite materials has been optimized.

2. Methods

In this section, the identities of the required materials and chemicals are revealed, while the synthesis of CuNPs and their impregnation on/into the OPFs is also specified. Thereafter, how the nanocomposites are prepared by using the CuNP impregnated OPFs and how they are characterized are also explained. After obtaining the nanocomposite materials, their properties (especially, mechanical strength and biodegradability) are analyzed as a function of the concentration of impregnated CuNPs. To optimize the process, and to obtain materials exhibiting the highest mechanical strength and controlled biodegradability, response surface methodology has been used. Therefore, the required instruments and the methods followed are highlighted under this section of the present article.

2.1. Materials and chemicals

Copper(II)chloride dihydrate salt (CuCl₂·2H₂O) of 98% purity and analytical grade sodium hydroxide were procured from Merck, Germany. Fully hydrolyzed PVA with molecular weights of M_w = 2000 and 70 000, sodium borohydride (NaBH₄; purity 99%), ascorbic acid (purity 99.7%) and a 60% solution of CHPTAC solution were purchased from Sigma-Aldrich. 2-Butanone peroxide (laboratory grade) was used as a cross-linker or hardener. Laboratory grade sodium hydroxide and acetone (Merck, Germany), were used and procured through Permula Chemicals Sdn Bhd., Malaysia. Raw OPFs were collected from the LKPP Corporation Sdn. Bhd., Kuantan, Malaysia. The UF was of golden-brown colour with an average diameter of 0.19 mm. Unsaturated polyester resin or virgin resin (UR) 268BQT was chosen as a composite matrix, which was procured from the local market through Permula Chemicals Sdn Bhd., Malaysia.

2.2. Preparation and impregnation of copper nanoparticles

The CuNP synthesis route has been followed from our own work and its characterization was also done previously.¹⁷ The cationization of fibres as well as the CuNP impregnation protocol was adopted from our other previous work¹⁸ and the details of this modification and characterization are discussed there. The loading of CuNPs (C_CuNPs) in the fibres was calculated according to the following eqn (1):

C_CuNPs = {(C_i − C_f) × 30} μg g⁻¹ (1)

where C_i and C_f are the concentrations (in ppm) of the copper sols before and after the adsorption of copper by the fibres.

2.3. Preparation of the composites

To prepare the composites, the UFs were chopped first to a length of ∼3 mm. Thereafter, the UFs were treated with a CHPTAC solution, followed by a CuNP sol treatment.¹⁸ During composite synthesis, such types of treated and untreated fibres of different percentages (0, 10, 20, 30 and 40% (v/v)) were used. UR was first mixed with 1 wt% 2-butanone peroxide (curing agent) for the matrix preparation of the composite. Then the required amount of chopped OPF or UF was mixed with the resin solution. The OPF–UR composites were prepared using a hand mixing technique. This technique was continued until complete mixing of resin and fibre was achieved. Thereafter, the respective resin–fibre mixture was subjected to a curing process at room temperature. Composite sheets were made using a 13 × 165 × 3 mm stainless steel mould, where polyvinyl alcohol (releasing agent) was coated onto the mould surface prior to the laying of the mixture. The mould was then covered by a plate and fixed by screwing the cover onto the base as tightly as possible. This covering of the mould is helpful to force the resin to penetrate into the fibre spaces. The mixture was left to cure for 24 h at room temperature.¹⁹ In the case of the composites, the volume fraction of the fibre (V_f) was calculated as per the following formula (eqn (2)):where w_f and w_m are the weight of the fibre and matrix respectively, and ρ_f and ρ_m are the density of the fibre and matrix respectively.

2.4. Characterization of the composites

FTIR spectra of the composites were recorded over the frequency range of 4000–650 cm⁻¹ using a Thermo Scientific Model Smart Performer ATR accessory with a Ge crystal, attached to a Thermo Scientific spectrophotometer (Model Nicolet Avatar-370) with a single bounce. The FTIR parameters were as follows: angle of incidence = 45°, sampling area = 2 mm; number of background scans = 32; number of scans = 32; optical resolution = 4.00 cm⁻¹.

2.5. Mechanical strength and biodegradability of the composites

Tensile tests were performed by using a Shimadzu (Model: AG-1) Universal tensile testing machine, in accordance with ASTM D638-77a standards. The specimens were tested at a rate of 5 mm per minute. The tensile strength was calculated from the stress–strain curve.

The biodegradability (BD) of the composites is correlated with the weight loss of the sample and this loss of weight is due to the composite samples being buried in soil. This test was performed using gardening soil and each specimen was buried in the soil and incubated at ambient temperature (25–30 °C). To maintain humid conditions, every 2 (two) days water was poured in the testing soil tray.²⁰ After burial for a duration of 90 days, the specimens were removed from the soil. Thereafter, the specimens were washed with water and dried at 40 °C in a vacuum oven up to a constant weight. The weight loss (%) due to biodegradation was calculated according to the following formula (eqn (3)):

BD (%) = {(w_bb − w_ab)/w_ab} × 100 (3)

where w_bb and w_ab are the dry weights of the composites before and after the burial, respectively.

2.6. Optimization using response surface methodology

The factorial design was developed using RSM based on the D-optimal design. This experimental design is a form of response surface methodology, and it is frequently using by researchers for performing numerous intentions such as design of experiments, analysis of variance, and empirical modelling. Compared with other response surface methods, D-optimal design has some advantages. Specifically, it involves a smaller number of experiments and can include categorical factors within the design of experiment system.²¹ Here, the D-optimal design has been performed for BD (in percentage of weight loss), TS and TM as the dependent variables (responses) while significant terms from the preliminary screening process were chosen as the independent factors. This design is quite popular, especially in looking for the maximum determinant of an information matrix (x′x), where x is the design matrix and this design can offer the best simultaneous prediction of the model parameters. More information about D-optimal design is available in the literature.^21–23

In this investigation, D-optimal design was conducted with a set of 12 composite designs of combination factors at two levels (high and low). The range and levels of the processing parameters involved with this design, along with the statistical results of the D-optimal plan are presented in Table 1, while the experimental design, layout and results (responses) of the D-optimal design are shown in Table 2. The analysis of variance for biodegradability is depicted in Table 3, where the independent variables are fibre loading (%) and amount of impregnated copper (μg g⁻¹), which are represented by the variables A and B, respectively. With the aim of modelling and optimization, a quadratic model is selected, although it is sometimes complicated to explain the important characteristics of the data using this model. The quadratic model for predicting the optimal point is in accordance with eqn (4)

Y = b₀ + ∑^​b_ix_i + ∑^​b_ii²x_i + ∑^​b_ijx_ix_j (4)

Where, Y is the response variable, b is the regression coefficient of the model, x is the coded levels of the independent variables, and i and j indicate the numeric factors. From eqn (4), it is observed that the model contains pure linear terms, two-term interactions and second order terms for quantitative factors.

Table 1 Variables of experimental design and statistical results of the D-optimal plan

Name	Units	Type	Std. dev.	Low	High
Fibre loading (x1)	%	Factor	0	20	40.00
Impregnated copper (x2)	μg g^−1	Factor	0	0	2590.00
TS	MPa	Response	2.56	45.8	60.42
TM	MPa	Response	51.94	890.45	1304.37
Biodegradability	%	Response	0.61	6.51	12.36

---

Statistics	Value
Maximum prediction variance (at a design point)	0.917
Average prediction variance	0.650
Condition number of coefficient matrix	2.247
Scaled D-optimality criterion	2.271
Determinant of (x′x)^−1	4.590 × 10^−5
Trace of (x′x)^−1	2.058

---

Table 2 Experimental design, layout and results (responses) of the D-optimal design

Std.	Run	Factor 1	Factor 2	Response 1	Response 2	Response 3
		A: fibre loading, x1 (%, v/v)	B: impregnated copper, x2 (μg g^−1)	TS (MPa)	TM (MPa)	Biodegradability (%)
1	6	20.00	2590.00	55.16	1188.59	7.52
2	4	40.00	2590.00	56.44	1237.67	9.89
3	3	20.00	1295.00	53.505	1089.53	8.49
4	1	30.00	2590.00	60.42	1304.37	6.51
5	8	20.00	0.00	45.8	890.45	15.99
6	11	40.00	0.00	51.7	1032.9	25.22
7	10	30.00	648.00	58.33	1278.74	8.74
8	7	40.00	1295.00	53.57	1120.48	10.35
9	12	25.00	1943.00	56.77	1232.33	8.04
10	5	20.00	0.00	51.85	990.46	14.46
11	2	40.00	0.00	48.74	1032.91	26.72
12	9	40.00	2590.00	51.43	1137.6	10.9

---

Table 3 ANOVA for biodegradability^a

Source	Sum of squares	df	Mean square	F-Value	Prob > F
a Values of “Prob > F” less than 0.0500 indicate model terms are significant.^24b Significant.c Not significant. In this case A, B, AB and B^2 are significant model terms.
Model	479.78	5	95.96	23.63	0.0007^b
A	58.35	1	58.35	14.37	0.0091^b
B	220.25	1	220.25	54.24	0.0003^b
AB	34.64	1	34.64	8.53	0.0266^b
A^2	15.81	1	15.81	3.89	0.0959^b
B^2	54.37	1	54.37	13.39	0.0106^b
Residual	24.37	6	4.06
Lack of fit	21.56	3	7.19	7.69	0.0512^c
Pure error	2.81	3	0.94
Cor total	504.15	11
Std. dev.	2.02		R^2		0.9517
Mean	12.74		Adjusted R^2		0.9114
Predicted R^2	0.7154		AdeqPrecision		13.932

---

In general, the main objective of RSM is to optimize the response (Y) based on the considered factors.^23,24 The Design Expert software 7.1.6 was used to develop the experimental design and optimize the regression equation (eqn (4)). The statistical significance of the model equation was determined by performing Fisher’s statistical test for analysis of variance (ANOVA).²³

3. Results and discussion

3.1. Characterization of the composites

Fig. 1 depicts the FTIR spectra of the virgin resin and the different fibre reinforced composites. In the spectrum of UR, the carbonyl (C=O) stretching (for ester linkage) is obvious at 1730 cm⁻¹ (ref. 3). However, in the composites, this peak is observed to be less prominent (compared with the same peak of UR) at 1711 cm⁻¹. Therefore, a 19 cm⁻¹ red shift is observed which indicates that in the composite of UF, the O–H groups (from fibres) form hydrogen bonds with the carbonyl groups (C=O) of UR. Besides, the spectra of the composites of UF and NF are more or less similar, except for the nature of the peak at around 1684 cm⁻¹ (C=O stretching), which is found to be due to the treatment of fibres. In the composite of NF, the ascribed absorption frequency of the C=O groups is more downfield than in the composite of UF, leading to a greater red shift of the C=O group peak observed for the composite of NF, which may confirm a stronger interaction between the C=O and O–H groups.^3,17,18 In addition, due to the treatment of the fibre, the non-cellulosic components are also removed, leading to an improvement in the crystallinity of cellulose, as well as an increase in the opportunity for hydrogen bonding interactions between C=O and O–H groups.²⁵ This finding is also consistent with the XRD pattern of fibres reported in our previous work.¹⁸ Thus, FTIR analysis confirms the formation of nanocomposite materials through the bonding interaction between the nanoparticle impregnated fibres and the unsaturated polyester resin.

The cationization of fibres as well as the CuNP impregnation protocol was adopted from our own work.¹⁸

Fig. 1 FTIR spectra of: (a) UR, (b) UF and (c) NF reinforced composites.

3.2. Biodegradability of the composites

Modelling outcomes. To study the effects and interactions of fibre loading (%) and impregnated CuNPs (μg g⁻¹) on biodegradability of composites, the obtained experimental data were analyzed by response surface methodology (Fig. 2). The RSM experimental design has been conducted systematically as an initial screening process, as summarized in Table 2. Under the RSM investigation, analysis of variance, confidence intervals, effect of factors on 3D-surface representations of responses, desirability functions, and optimization details have been focused on.

Fig. 2 Normal plot of residual for the response biodegradability of nanocomposites.

The relative effect of impregnated copper as a working parameter on the biodegradability was obvious since the experiments were performed at low loading levels (0 to 647.5 μg g⁻¹; e.g., runs 2, 5, 8, 10 and 11), and a high percentage of biodegradability was demonstrated. Such a pattern in the biodegradability of materials is expected in some cases. On the other hand, at the highest concentration of impregnated copper (2590.00 μg g⁻¹; e.g., runs 1 and 6) a low percentage of biodegradability was experienced. This low percentage of biodegradability is also obvious in some other cases. This finding indicates that due to an increase in the dose of loaded CuNPs per g of fibre, the biodegradability was controlled, potentially leading to a high restriction of micro-organism attack on the fibres of the composite. This fact is rational, as Cu nanoparticles are one of the long-lasting biocides and have the potential ability for sterilization to a good extent.¹⁸ This sterilization effect originates from the catalytic properties of Cu nanoparticles, that help to produce active oxygen in water, which dissolves organic substances and ultimately can cause sterilization to a certain level.^26,27

Analysis of variance. ANOVA was performed to analyze the results statistically. Based on the experimental results, variance analysis was performed, where a quadratic model is suggested for biodegradability. Some statistics about the statistical validity of the models are shown in Table 3. From the obtained statistical indicators (as shown in Table 3) the following remarks may be true:

1. The model F-value of 23.63 implies that the model is significant. There is only a 0.07% chance that a “Model F-Value” this large could occur due to noise.

2. Values of “Prob > F” less than 0.0500 indicate model terms are significant. In this case A, B, AB and B² are significant model terms. It is considered that values greater than 0.1000 indicate the model terms are not significant.

3. The “Lack of Fit F-value” of 7.69 implies that there is a 5.12% chance that a “Lack of Fit F-value” this large could occur due to noise. A not significant lack of fit is good – we want the model to fit.

4. The predicted R² is in reasonable agreement with the adjusted R² (accounting for the number of predictors in the models). This finding indicates that in the prediction, the power of the model and the fitting of the data is good.

5. “Adeq Precision” measures the signal to noise ratio. A ratio greater than 4 is desirable in the analysis and the obtained ratio of 13.932 indicates an adequate signal. This model can be used to navigate the design space.

By using the experimental data of fibre loading (A) and impregnated copper (B) (shown in Table 3) as the two main variables, the multiple regression, eqn (5), for biodegradability is presented below:

Biodegradability = +6.64 + 2.59A − 5.31B − 2.29AB + 3.07A² + 5.31B² (5)

ANOVA was performed in order to statistically analyze the results. Significant process parameters were identified, and interaction effects of process parameters were also studied.

The well-known statistical analysis using Fisher’s statistical test analysis of variance, i.e. ANOVA, was performed to determine the significant variables where the degree of significance was ranked on the basis of the F-ratio value.²⁴ Indeed, a greater “F-value” corresponds to a smaller “Prob > F”, leading to the greater significance of the corresponding model and the individual coefficients.^23,28 The results of the analysis (as depicted in Table 3) confirm that the confidence level was greater than 95% (p < 0.05) for the biodegradability performance, while the F-value and P-value of the model were 20.61 and 0.001, respectively. This finding indicates that the estimated mathematical model is significantly fitted with the experimental data.

From Table 3, it is evident that factor A, i.e. the percentage of fibre loading, is directly proportional to the response 3, i.e. the biodegradability of the composites, though one exception is observed. At a low percentage of fibre loading (20%, e.g., runs 3, 5, 6 and 8), the biodegradability of the composites was also low. The exact opposite scenario was observed for a high percentage of fibre loading (40%, e.g., runs 2, 4, 7, 9 and 11). From these facts, it can be considered that with an increase in the fibre loading, the density of the fibres in the composites is increased, leading to a lack of wet-ability of the fibres, which increases the possibility of the degradation of the fibres and ultimately the composites also. Comparing runs 6 (20%, 2590 μg g⁻¹), 4 (40%, 2590 μg g⁻¹) and 1 (30%, 2590 μg g⁻¹), the biodegradability is 7.52, 9.89 and 6.51% respectively as shown in Table 2. Here, the 30% fibre loaded composite has the lowest biodegradability (with the same amount of impregnated copper). It can be considered that the property of biodegradability is inversely proportional to the ultimate stress of any composite, which depends on several factors. Amongst these, the major role playing factors are the properties of the reinforcement and matrix, interfacial adhesion and the fibre volume fraction.^29–32 The fibre mechanical properties, such as the initial modulus and ultimate tensile stress, are related not only to the chemical composition of the fibre but also to its internal structure. However, a considerable scattering of the measured tensile strength was experienced by other researchers, which may happen due to the different degree of adhesion between fibre and matrix.³³ The observed minimum biodegradability in the 30% fibre loaded composite may be due to the ultimate cellulose content of the individual fibres, low microfibrillar-angle (i.e., the angle between the fibre axis and the fibril of the fibre)³¹ and the interfacial adhesion. Strong adhesion between the fibres and the matrix can cause a decrease in water absorption, leading to a decrease in the biodegradability.³⁴ It was also reported that fibrillation can increase the available effective surface area in contact with the matrix, leading to an improvement in interfacial bonding.²⁸⁻³¹ Therefore, in this study (for TS and BD) it can be concluded that the optimum fibre loading was 30 vol% among all of the reinforced composites and a similar trend is observed elsewhere.³⁵

The confidence interval results are shown in Table 4, where the standard error of the process parameters for biodegradability is within the limits i.e. ∼5%.

Table 4 Confidence interval^a

Factor	Coefficient estimate	df	Standard error	95% CI low	95% CI high	VIF
a Where, df – degree of freedom, CI – confidence interval, and VIF – variance inflation factor.
Intercept	6.64	1	1.49	2.99	10.29
A-fibre loading	2.59	1	0.68	0.92	4.26	1.06
B-impregnated copper	−5.31	1	0.72	−7.08	−3.55	1.09
AB	−2.29	1	0.78	−4.20	−0.37	1.06
A^2	3.07	1	1.56	−0.74	6.89	1.16
B^2	5.31	1	1.45	1.76	8.85	1.09

---

A normal plot of residuals displays the residual data where 98% of residuals should fall within 3 sigma, and in this experimental design the data is well within three 3 sigma.

The effect of varying the fibre loading and impregnated copper content on the biodegradability was further analyzed using a simulated three dimensional response surface and contour plots. The effect of varying the fibre loading and impregnated copper content on the biodegradability in a 3D-representation of the responses (with the considered two factors) is depicted in Fig. 3. This figure demonstrates that the biodegradability increases when the fibre loading (%) is changed from 20.0 to 40.0 and as the impregnated copper quantity is decreased from 2590 μg to 0. Based on these results, the minimum biodegradation of 6.51% is obtained when the fibre loading percentage and the quantity of copper loading is 30% and 2590 μg, respectively. It is also possible to state that the loaded CuNPs provide the principle effect on the biodegradability. In addition, fibre loading is considered as a secondary factor.

Fig. 3 Effect of factors on 3D-surface representations of responses.

Among the 12 runs, from the runs 1, 2, 4, 6 and 11 it can be concluded that factor 2 (the impregnated copper) is the predominating factor on the response of biodegradability. Furthermore, it is also stated that the main effect of fibre loading (A), impregnated copper (B) and the second order effects of B² and two level interactions of AB were significant as model terms (factors). However, other model terms, especially the second order term (A²), were relatively less significant as their confidence level was below 95% (p > 0.05). Besides, the “Lack of Fit” value was found (Prob > F = 0.0512) to be not significant, which meant that the model was suitably fit. Therefore in this analysis, for significant biodegradability, the model term rankings are as follows: B < A < B² < AB. Moreover, the coefficient of determination (R²) of the model was 0.95 ≈ 1 (which is supposed to be 1), indicating a good correlation (about 95% of the variability of the data) between the experimental and predicted values. The “Predicted R²” of 0.9517 is in reasonable agreement with the “Adjusted R²” of 0.9114. Moreover, “Adeq Precision” measured a good signal to noise ratio.

Desirability function. During optimization, the function of desirability analysis in RSM was also considered and effectively analysed. Fig. 4 shows the composite desirability of the biodegradability. The plot and the value of the composite desirability were obtained using the design expert software. Generally, composite desirability (d_G) is measured using the individual desirability index of all the responses, by combining them to form a single value called the composite desirability (d_G) by using the following eqn (6):²⁴where d_i is the individual desirability of the property, w_i is the weight of the property and w is the sum of the individual weights. It is known that a prediction with a high desirability is ideal, and a high desirability value (1.00) is an indication of better product quality, which is expected to predict the values of the process parameters.

Fig. 4 The desirability of the composites (with B (in μg g⁻¹) and A (in vol%)).

3.3. Validation of the optimization process

In order to confirm the prediction accuracy and to show the acceptability of the developed empirical model, it is necessary to validate the proposed optimization parameters. This is performed using the regression equation for predicting the biodegradability at any particular factor A and B within the range of the levels defined previously. To validate the optimization parameters, experimental rechecking was performed using the nearest proposed conditions that were not done previously, and the obtained results from the prediction and experiment (based on proposed conditions) are shown in Table 5. The obtained experimental value and the corresponding predicted value were compared and the percentage of error was evaluated. The percentage of error between the experimental and predicted value of the response over a selected range of operating levels is calculated as per eqn (7).

Table 5 Validation of the model

Fibre loading (%)	Impregnated copper (μg g^−1)	Biodegradability (%)		Error (%)
		Predicted	Experimental
30.00	2532	6.39	6.19	−3.23
35.00	99	9.40	9.75	+3.72

---

Table 5 reveals that the percentage errors are 3.23 and 3.72 (both are below 5%) and the negative sign means that the experimental value is less than the predicted value. These acceptable percentage errors between the experimental and predicted values indicate that the optimization is reasonably adequate and can satisfy to within 95% of the prediction interval. Therefore, it is possible to say that the predicted mathematical optimization is noticeably accurate for responding to the term of interest (biodegradability).

It was mentioned that to analyze the response biodegradability of the composites, two factors were considered and it was also stated that the impregnated copper was the dominating factor. Nevertheless, simply disregarding any factor is not wise, because in the responses all the individual factors that are involved in the interactions have to be considered jointly. Therefore during optimization, the targeted response (biodegradability) was set at the minimum and simultaneously other responses (both TS and TM) were set at the maximum. Not only that, optimization was also done by setting the targeted response (biodegradability) and mechanical strengths (both TS and TM) at the maximum. Thus, two models are predicted and the aim of the work to control the biodegradability of the materials has been addressed. Based on this analysis method, the predicted result of the biodegradability and also the experimental result of it, are shown in Table 5. To ensure a minimum biodegradability, the optimum fibre loading and the impregnated copper was 30.00% and 2532 μg g⁻¹, respectively. Conversely, the maximum biodegradability was obtained when the optimum fibre loading and the impregnated copper was 35.00% and 99 μg g⁻¹, respectively. It is also observed that the obtained experimental results of biodegradability are in good consistency with the predicted results with a relatively insignificant degree of error.

4. Conclusion

Globally, with the advancement of industrialization and the economy, the development of controlled biodegradable nanocomposites or sustainable materials is in demand. In this investigation, a CuNP impregnated OPF fibre reinforced nanocomposite was developed. Thereafter, it was characterized, and its mechanical strength and the control of the material’s biodegradability as a function of impregnated CuNP concentration were analyzed. The developed nanocomposites could be considered as sustainable materials as they are prepared from renewable resources and represent an important contribution to this field. FTIR analysis confirmed the bonding interaction between the nanoparticle impregnated fibre and the matrix. For the first time, the control of the biodegradability of a composite with the impregnation of nanoparticles is explained from the experimental findings along with the process of optimization. A maximum biodegradability of the nanocomposite is achieved when the fibre loading and the impregnated copper is 35.00% and 99.00 μg g⁻¹, respectively. Strong and durable nanocomposites have been achieved when the fibre loading and the impregnated copper is 30% and 2590 μg g⁻¹, respectively. The described work has established that for controlling the biodegradability of the material, the quantity of impregnated copper is a potential tool. The connection between the responses and variables used in this study is well explained by the predicted models. Moreover, the model terms were well explained and the prediction was also done successfully. Consequently, this work will be a very good guideline for the future, to control the biodegradability of materials, taking into account the specific requirements of their applications. The authors of this paper hope that the insights presented here will be of good value to future researchers, who have the aim to develop such types of controlled biodegradable materials or sustainable materials at the pilot scale and up to an industrial level. In the future, if any work is performed regarding the control of biodegradability, it will be interesting to investigate if our model can still predict the results if different copper nanoparticles are impregnated (say using a different size of nanoparticles), and can still predict the biodegradability well when the fibre loading is not in the range of 20–40% or the impregnated copper concentration is not in the range of 0–2590 μg g⁻¹.

Abbreviations

CHPTAC	(3-Chloro-2-hydroxypropyl) trimethylammonium chloride
ANOVA	Analysis of variance
AMTEC	Advanced membrane technology research center
ASTM	American society for testing and materials
ATR	Attenuated total reflectance
BD	Biodegradability
CuNPs	Copper nanoparticles
FTIR	Fourier transform infrared spectroscopy
OPF	Oil palm fibre
ppm	Parts per million
NPs	Nanoparticles
NF	CuNP impregnated fibre
NF–C	NF reinforced composite
PVA	Poly(vinyl alcohol)
RSM	Response surface methodology
TM	Tensile modulus
TS	Tensile strength
ISO	The international organization for standardization
UR	Unsaturated polyester resin
UF	Untreated OPF
UFs	Untreated fibres
XRD	X-Ray diffraction

Acknowledgements

The authors are grateful to Dr Rodrigo Lozano, professor at Utrecht University, Utrecht, The Netherlands, who provided expert advice for improving the manuscript. The authors would also like to thank University Malaysia Pahang and University Technology Malaysia for funding (GRS110322) and the required support to complete this study.

References

1. R. Lozano and D. Huisingh, J. Cleaner Prod., 2011, 19, 99–107.
2. R. Lozano, J. Cleaner Prod., 2008, 16, 1838–1846.
3. M. N. K. Chowdhury, A. F. Ismail, M. R. Khan, M. D. H. Beg, M. H. D. Othman, R. J. Gohari and A. Moslehyani, RSC Adv., 2015, 5, 100943–100955.
4. D. Lilley and V. Lofthouse, Eng. Educ., 2009, 4, 29–41.
5. B. Roy, P. Bharali, B. Konwar and N. Karak, Bioresour. Technol., 2013, 127, 175–180.
6. J. M. Garces, D. J. Moll, J. Bicerano, R. Fibiger and D. G. McLeod, Adv. Mater., 2000, 12, 1835–1839.
7. J. Holbery and D. Houston, JOM, 2006, 58, 80–86.
8. F. Hussain, M. Hojjati, M. Okamoto and R. E. Gorga, J. Compos. Mater., 2006, 40, 1511–1575.
9. F. Gao, Mater. Today, 2004, 7, 50–55.
10. M. Alexandre and P. Dubois, Mater. Sci. Eng., R, 2000, 28, 1–63.
11. P. Mangalgiri, Bull. Mater. Sci., 1999, 22, 657–664.
12. M. Chowdhury, M. Beg and M. R. Khan, Procedia Eng., 2013, 68, 431–438.
13. V. Rossi, N. Cleeve-Edwards, L. Lundquist, U. Schenker, C. Dubois, S. Humbert and O. Jolliet, J. Cleaner Prod., 2015, 86, 132–145.
14. N. Larson and H. Ghandehari, Chem. Mater., 2012, 24, 840–853.
15. H. Tian, Z. Tang, X. Zhuang, X. Chen and X. Jing, Prog. Polym. Sci., 2012, 37, 237–280.
16. Y. Ikada and H. Tsuji, Macromol. Rapid Commun., 2000, 21, 117–132.
17. M. Chowdhury, M. Beg, M. R. Khan and M. Mina, Mater. Lett., 2013, 98, 26–29.
18. M. Chowdhury, M. Beg, M. R. Khan and M. Mina, Cellulose, 2013, 20, 1477–1490.
19. R. Ojah and S. Dolui, Bioresour. Technol., 2006, 97, 1529–1535.
20. N. Teramoto, K. Urata, K. Ozawa and M. Shibata, Polym. Degrad. Stab., 2004, 86, 401–409.
21. E. Kuram, B. Ozcelik, M. Bayramoglu, E. Demirbas and B. T. Simsek, J. Cleaner Prod., 2013, 42, 159–166.
22. R. H. Myers, D. C. Montgomery and C. M. Anderson-Cook, Response surface methodology: process and product optimization using designed experiments, John Wiley & Sons, 2009.
23. A. Zularisam, A. F. Ismail, M. Salim, M. Sakinah and O. Hiroaki, J. Membr. Sci., 2007, 299, 97–113.
24. R. K. Bhushan, J. Cleaner Prod., 2013, 39, 242–254.
25. A. Moshiul Alam, M. Beg, D. Reddy Prasad, M. Khan and M. Mina, Composites, Part A, 2012, 43, 1921–1929.
26. N. C. Cady, J. L. Behnke and A. D. Strickland, Adv. Funct. Mater., 2011, 21, 2506–2514.
27. D. Chattopadhyay and B. Patel, Indian J. Fibre Text. Res., 2009, 34, 368.
28. D. C. Montgomery, Design and analysis of experiments, John Wiley & Sons, 2008.
29. A. Peled, Workshop on high performance fiber reinforced cement composites (RILEM) HPFRCC-5, Mainz, Germany, 2007.
30. A. A. Abdulmajeed, T. O. Närhi, P. K. Vallittu and L. V. Lassila, Dent. Mater., 2011, 27, 313–321.
31. P. Wambua, J. Ivens and I. Verpoest, Compos. Sci. Technol., 2003, 63, 1259–1264.
32. M. Baiardo, E. Zini and M. Scandola, Composites, Part A, 2004, 35, 703–710.
33. S. Shibata, Y. Cao and I. Fukumoto, Polym. Test., 2005, 24, 1005–1011.
34. V. Vilay, M. Mariatti, R. Mat Taib and M. Todo, Compos. Sci. Technol., 2008, 68, 631–638.
35. P. Sreekumar, K. Joseph, G. Unnikrishnan and S. Thomas, Compos. Sci. Technol., 2007, 67, 453–461.

---