Extraction and Characterization of Microcrystalline Cellulose (MCC) from Durian Rind for Biocomposite Application

Abstract

The scientific community has become interested in cellulose extraction from agro-industrial wastes because they contain large quantities of cellulose and are being researched globally due to their application. Cellulose extracted from these wastes offers varying characteristics and structures depending on the specific source. In this study, microcrystalline cellulose (MCC) was extracted from different parts of the durian husk through acid hydrolysis with various concentrations of oxalic acid and cellulose-to-acid ratios. The durian husk underwent a pretreatment method involving bleaching and alkaline treatment. The characteristics of MCC, such as functional group, morphology, and crystallinity, were studied. The findings indicated that lignin was successfully removed through pretreatment. However, traces of hemicellulose were still detected in a few samples. The extracted MCCs demonstrated a diameter range from 6.237 to 25.38 μm and crystallinity within the range of 71.43–78.30%. Polycaprolactone (PCL)-based biocomposites with different MCC weights (0.2, 0.5, and 1.0 wt%) were fabricated, and the performance was evaluated through tensile testing and biodegradability tests. The addition of 1.0 wt% MCC enhanced the tensile strength by 22%. SEM analysis revealed the cross-section fracture surface of the biocomposite, indicating the contribution of the MCC during pull-out. Meanwhile, the biodegradability of the biocomposite increased with the addition of MCC, indicating that the MCC-based biocomposite was fully biodegradable.

Introduction

The adverse impacts of population growth on our natural environment are increasingly evident and have imposed significant stress on the Earth’s ecosystems. The production of waste is one of the most pressing issues. Around the world, one-third of the food produced is wasted, producing 1.3 billion tons of biowaste [1]. Most of the biomass waste is generated from the agricultural-food industry. The growing population and increasing global consumption of plastic materials have also led to significant plastic waste issues. Eight to nine million tons of plastic, mainly from single-use items, end up in the ocean each year [2]. As a result, one million seabirds and 100 million marine species die due to the plastic in the ocean [3]. At this rate, more plastic might be in our ocean than fish by 2050 [4]. Correspondingly, the development of novel sustainable polymers created using natural origin is of growing interest due to environmental issues and consumer requirements for higher quality and longer shelf life of food [5].

Cellulose is the primary component of all plant materials and is consistently regenerated by photosynthesis. About 1/2 to 1/3 of all plant tissue comprises cellulose, which confers stability and strength to the plant cell walls [6]. Cellulose present in plant material can be divided into amorphous and crystalline regions. Cellulose is predominantly composed of the crystalline structure, while the amorphous region makes up a relatively minor proportion [7]. Applying suitable treatments, such as enzymes, chemicals, or mechanical processes, to lignocellulosic biomass can effectively isolate cellulose with a crystalline structure, producing microcrystalline cellulose (MCC) [8]. MCC can subsequently be utilized as a source to produce cellulose nanocrystals (CNC).

MCC is a finely textured, odorless, and white substance, presenting itself as a crystalline powder with excellent mechanical strength, low density, biocompatibility, non-toxicity, reusability, and biodegradability [9]. These characteristics render MCC highly desirable as the reinforcing filler in polymer blends over alternative fibers [10]. It is also recognized as the most robust and rigid natural reinforcing filler, primarily attributed to its large surface area compared to other typical cellulose fibers [11]. The readily available MCC in markets is predominantly derived from virgin sources, such as purified cotton or expensive hardwood. As a result, current research efforts have shifted toward discovering various underutilized lignocellulosic materials biomass waste as inexpensive, diverse, and sustainable sources of MCC. This research work aims to produce MCC by converting waste by-products into value-added products [12].

Durian, known as the “king of fruits”, scientifically referred to as Durio zibethius, belongs to the family Bombacaceae and genus Durio. It is a tropical fruit with a skin full of sharp thorns. Durian rind, also known as durian husk, makes up over 50% of the durian’s quality, and the inner white part of the durian husk is referred to as the white sac [13]. Durian is one of the most popular tropical fruits in Southeast Asia, including Malaysia, Philippines, Indonesia, and Thailand [14]. Collectively, these countries produce over 1,600,000 tons of durian each year.

Moreover, the global durian market, valued at USD 25.01 billion in 2023, is expected to grow substantially with a projected compound annual growth rate of 7% [15]. The rapid urbanization in emerging economies causing the rise of disposable incomes has fueled an expansion of the middle-class demographic. As a result, there is increased demand for premium food products, such as durian pastries and desserts. However, the rise in durian food processing, including the production of durian pulp and paste, has led to an accumulation of non-consumable durian husk waste, presenting potential environmental challenges that must be addressed [16]. One way to tackle this problem is by transforming durian husk waste into valuable resources. Durian husk is a lignocellulosic biomass comprising three primary components of approximately 30.7% hemicellulose, 15.6% lignin, and ranging from 57 to 64% cellulose [14]. The high cellulose content found in durian husk presents an opportunity to extract valuable MCC. The common method used in most research to extract MCC is acid hydrolysis using sulfuric acid. However, sulfuric acid is a strong acid and has several significant disadvantages, such as being hazardous, corrosive, and having environmental compatibility issues [17]. Consequently, exploring alternative milder acids, such as oxalic acid, for MCC extraction is advisable. While there is limited research using oxalic acid for MCC extraction from durian husk, determining the appropriate acid concentration, temperature, and cellulose-to-acid ratio is a task to be undertaken in this study.

In response to the global plastic waste problem, researchers and industries are increasingly exploring the potential of biopolymers or biodegradable polymers, which are derived from living things such as plants and microorganisms. These polymers can be found naturally in human bodies or chemically synthesized from vegetable oils, fats, proteins, or amino acids. This distinction between synthetic and natural polymers arises from their production methods and varied sources [18]. In recent years, synthetic biopolymers have garnered attention because of their notable benefits in flexibility and stability that suit a wide range of applications. Polycaprolactone (PCL), polyglycolic acid (PGA), polydioxanone (PDS), polylactic acid (PLA), and polypropylene fumarate (PPF) are some examples of synthetic biodegradable polymers [19]. Despite the eco-friendly nature of biopolymers, their inherent characteristics typically result in lower mechanical strength and stiffness compared to conventional plastics. However, previous research works have demonstrated the extraction of microcrystalline cellulose from biomass, showcasing its potential as a promising reinforcing agent for biopolymers [20; 21]. Despite the advancements in MCC extraction, previous studies have primarily relied on strong acids, raising concerns about safety and environmental impact. Additionally, many of these studies focused solely on a single set of operating parameters on acid hydrolysis, leaving a knowledge gap in understanding the effects of various conditions on MCC extraction and its performance as a reinforcing agent.

Hence, the present study aims to investigate the feasibility of durian rind to produce microcrystalline cellulose (MCC) as a natural reinforcing agent for biocomposite applications. MCC was extracted from durian rind through a series of chemical pre-treatments followed by oxalic acid hydrolysis at different acid concentrations and cellulose-to-acid ratios. The effectiveness of MCC for biocomposite applications was tested by incorporating it with PCL, evaluating its mechanical properties, biodegradability, and thermal stability.

Methodology

Materials

The durian husk was obtained from a local fruit farm in Kuching, Sarawak. Sodium hydroxide (Quality Reagent Chemical), aqueous glacial acetic acid (HmbG Chemicals), sodium chloride (Quality Reagent Chemicals), sodium hypochlorite (Bendosen), oxalic acid (Merck), PCL (poly(3-caprolactone), and tert-butanol (Merck) were used in this study.

Experimental Flowchart

The MCC extraction process in this study (Fig. 1) starts with an alkaline treatment to remove hemicellulose from the durian rind, followed by a bleaching treatment to remove the lignin. To study the effect of a bleaching agent on the removal of lignin, two routes were applied: the combination of sodium chloride (NaCl) and acetic acid (CH₃COOH), and sodium hypochlorite (NaClO). The bleached cellulose then underwent acid hydrolysis with varying conditions to study the yield and purity of the extracted MCC. The cellulose-to-acid ratio is one of the crucial factors for MCC extraction; hence, ratios of 1:10 and 1:50 were used in this study. Additionally, different oxalic acid concentrations [30, 50, and 65% (w/v)] were also examined to evaluate the impact of acid hydrolysis on MCC properties. The extracted MCC was characterized through various analyses. Subsequently, the MCC was mixed with polycaprolactone (PCL) to fabricate biocomposites and further evaluated their mechanical properties, biodegradability, and thermal stability.

Fig. 1

Flowchart of the experimental procedure and the designated parameters used in this study

Durian Rind Moisture Content

Different parts of durian rind, i.e., durian husk (DH) and white sac (WS), were cut into smaller pieces and washed thoroughly under running water to remove contaminants. The initial mass of DH and WS was weighed and recorded as D₁ and dried for 24 h at 100 ºC in a Memmert oven [22]. The final mass of the dried DH and WS were recorded and designated as D₂. Equation (1) was used to calculate the moisture content and Fig. 2 shows the DH and WS pieces before and after drying.

$$\:\text{M}\text{o}\text{i}\text{s}\text{t}\text{u}\text{r}\text{e}\:\text{c}\text{o}\text{n}\text{t}\text{e}\text{n}\text{t}\:\left(\text{\%}\right)=\frac{{\text{D}}_{1}-{\text{D}}_{2}}{{\text{D}}_{1}}\times\:100\text{\%}$$

(1)

Fig. 2

Durian pieces of (a) DH (bottom) and WS (top) before drying, and (b) DH (left) and WS (right) after drying

Preparation of Raw Materials

To extract the cellulose, DH and WS were ground using a Fritsch rotor mill. The grinding process for the rotor mill involves using an 8-rib rotor and a 1.0 mm Trapez sieve ring made of special steel to ensure the particle size is 1.0 mm. The ground DH and WS powder (Fig. 3) were kept in a desiccator until further use to prevent exposure to moisture in the air [23].

Fig. 3

Dried durian husk (DH) and white sac (WS) powder

Alkaline Treatment

Kim et al. [23] reported that the ideal conditions of alkaline pretreatment to remove non-cellulosic materials from biomass are using 0.5–10% (w/v) of sodium hydroxide (NaOH), under a temperature of 60–180 ℃, and 0.5–10% of solid loading. In this study, 10 g of the DH and WS was treated with 100 mL of 2% (w/v) NaOH solution at a temperature between 90 and 100 ℃ for 3 h with continued stirring on the IKA hot plate stirrer. The mixture was then cooled to room temperature (25 ℃) and filtered through distilled water. The pore size of the filter paper used was 125 mm in diameter. After filtering the residue, the filtrate was repeatedly rinsed with distilled water until the pH reached neutral to eliminate the alkali. The resultant residue was dried in the oven for 24 h at 60 ºC. The obtained alkaline-treated DH and WS were then ground using a blender and kept in a desiccator until further use.

Bleaching Treatment

The bleaching process was conducted by mixing the alkaline-treated cellulose with the bleaching agent. Two bleaching agents were used to study the effect of the bleaching agent on further removal of lignin from alkaline-treated cellulose. The first bleaching agent was the combination of 0.50% (v/v) aqueous glacial acetic acid (CH₃COOH) and 1.50% (w/v) sodium chloride (NaCl), while the second bleaching agent was 5% (v/v) of sodium hypochlorite (NaClO).

The bleaching process was carried out at 70 ºC for 2 h with constant stirring. The bleaching solution was then cooled to room temperature, and the residue was filtered and washed multiple times with distilled water to obtain a neutral pH residue. The bleached cellulose was dried at 60 °C in an oven for 24 h, followed by grinding in a rotor mill. Figure 4 shows the extracted cellulose using different types of bleaching agents. According to Rahman et al. [22], the yield of the extracted cellulose from DH and WS can be calculated using Eq. (2):

$$\:\text{\%}\:\text{Y}\text{i}\text{e}\text{l}\text{d}\:\text{o}\text{f}\:\text{C}\text{e}\text{l}\text{l}\text{u}\text{l}\text{o}\text{s}\text{e}=\frac{{\text{W}}_{\text{Alkaline-}\text{treated}\:\text{Cellulose}}\:\left(\text{g}\right)}{{\text{W}}_{\text{D}\text{H}/\text{W}\text{S}\:}\left(\text{g}\right)}\times\:100\text{\%}$$

(2)

Fig. 4

Cellulose treated with (a) 1.50% (w/v) NaCl and 0.50% (v/v) CH₃COOH, and (b) 5% (v/v) NaClO

Where $\:{\text{W}}_{\text{C}\text{e}\text{l}\text{l}\text{u}\text{l}\text{o}\text{s}\text{e}}$ is the weight of extracted cellulose and $\:{\text{W}}_{\text{D}\text{H}/\text{W}\text{S}\:}$ is the weight of DH or WS.

Microcrystalline Cellulose (MCC) Extraction

To extract crystalline cellulose, oxalic acid was used in this study. Oxalic acid is a weak organic acid that requires a higher temperature (above 80 ℃) for efficient production of crystalline cellulose [24]. Raising the hydrolysis temperature promotes the defibrillation of the feed fiber by reducing the degree of polymerization of cellulose fiber which is beneficial for producing crystalline cellulose. On the other hand, if the reaction time is less than 4 h, the hydrolysis process might not adequately break down the cellulose fiber to extract crystalline cellulose efficiently [25]. Hence, extending the reaction time could boost extraction yield. Nevertheless, it is important to note that prolonging the reaction time may also lead to excessive hydrolysis, ultimately leading to a reduced extraction yield [25]. A study by Fitriani et al. [26] examined the duration of acid hydrolysis on crystalline cellulose properties and found that 3 h is efficient for producing cellulose nanocrystals. The crystalline cellulose underwent 3 h of acid hydrolysis exhibit the highest crystallinity and smallest particle size, and produced a stable aqueous suspension of crystalline cellulose. Therefore, the following hydrolysis settings are established.

Solution with oxalic acid concentrations of 30, 50 and 65% (w/v) were weighed and mixed with distilled water in a 250 mL beaker, respectively. The mixture was then heated on the hot plate with continuously stirring until the oxalic acid was fully dissolved in the distilled water. This is because dehydrated oxalic acid cannot dissolve in distilled water at room temperature. Two different cellulose-to-oxalic acid ratios, 1:50 and 1:10, were used to study the effect of the raw material-to-acid ratio on the extraction yield. Table 1 shows the list of samples. The bleached cellulose and oxalic acid mixture was poured into a test tube and placed inside the water bath, then heated to a temperature of 90 ºC for 3 h. The hydrolysis process was stopped by adding a 10-fold excess of distilled water to the reaction mixture [27]. The resulting mixture was instantly placed in the centrifuge tubes and centrifuged for 10 min at 6000 rpm to separate the MCC and the distilled water [28].

Following centrifugation, the liquid supernatant in the centrifuge tubes was drained, and the remaining MCC suspension was transferred to a dialysis tube. The dialysis tube, filled with the MCC suspension, was immersed in a beaker containing distilled water. The distilled water in the beaker was replaced daily until its pH became neutral. To minimize the suspension agglomeration after acid hydrolysis, the MCC suspension was subjected to 10 min of sonication at 25 ºC using an ultrasonic probe [29].

Subsequently, tert-butanol (t-BuOH) was added as a surfactant to the colloidal suspension before drying the MCC suspension in the oven. The co-solvent was prepared by combining t-BuOH and suspension in a ratio of 1:9, as proposed by Hanif et al. [30]. The mixture was stirred until a homogenous dispersion was achieved. After that, the mixture was poured into a beaker and dried in an oven at 100 ºC for 24 h. Figure 5 illustrates the extracted MCC.

Fig. 5

MCC extracted from DH or WS

The yield percentage of extracted MCC can be calculated as given by Eq. (3):

$$\:\text{\%}\:\text{Y}\text{i}\text{e}\text{l}\text{d}\:\text{o}\text{f}\:\text{M}\text{C}\text{C}=\frac{{\text{W}}_{\text{M}\text{C}\text{C}}\:\left(\text{g}\right)}{{\text{W}}_{\text{c}\text{e}\text{l}\text{l}\text{u}\text{l}\text{o}\text{s}\text{e}\:}\left(\text{g}\right)}\times\:100\text{\%}$$

(3)

where $\:{\text{W}}_{\text{M}\text{C}\text{C}}$ is the weight of MCC and $\:{\text{W}}_{\text{c}\text{e}\text{l}\text{l}\text{u}\text{l}\text{o}\text{s}\text{e}\:}$ is the weight of extracted cellulose.

Table 1 List of MCC samples

Fourier Transform Infrared (FTIR) Analysis

The functional group changes in the chemical structure of lignin, cellulose, and hemicellulose of the extracted MCC were characterized using a Perkin Elmer Spectrum 3 FTIR spectrometer. The oven-dried MCC samples underwent 16 scans at a resolution of 4 cm⁻¹, over the range of 450 cm⁻¹ to 4000 cm⁻¹ [31].

X-Ray Diffraction (XRD) Analysis

The crystallinity index of the extracted MCC was determined using a PANalytical (X’pert Pro, Netherlands) X-ray diffractometer with Cu K_α radiation (λ = 1.54060Å). The MCC was scanned at a step increment of 0.02° within 2θ range of 7° to 60° with X-ray tube voltage and current of 45 kV and 35 mA, respectively. Equation (4), which uses the Segal approach, was used to estimate the relative crystallinity index of MCC, Crl (%).

$$\:\text{C}\text{r}\text{l}=\left[\frac{I-{I}_{am}}{I}\right]\times\:100\text{\%}$$

(4)

where $\:I\:$ represents the peak intensity corresponding to crystalline cellulose, while $\:{I}_{am}$ is the peak intensity of the amorphous region.

Ultraviolet-Visible (UV-vis) Analysis

To determine the turbidity of the suspensions, which correlates with the size of MCC, UV-vis spectroscopy was performed using a Perkin Elmer Lambda 35 at room temperature ranging from 200 to 800 nm in transmittance mode [32]. The extracted MCC was mixed with distilled water, then transferred into a cuvette and filled to ¾ with the solution.

Morphological Analysis

Field emission scanning electron microscope (FESEM) (Quanta 400, FEI, Switzerland) was employed to examine the surface morphology of the MCC. The FESEM analysis was conducted using a 2–10 kV acceleration voltage and a 6–10 mm working distance. To prevent charging and ensure high-quality images, the samples were sputter-coated with platinum before analysis [31]. In addition, the MCC particle size distribution was determined using a Malvern Zetasizer Nano ZS.

Fabrication of Biocomposite

To fabricate the PCL/MCC biocomposite specimens, the PCL granules were first melted at 70 ºC in a water bath. MCC powder was gradually added to the molten PCL (Fig. 6a) with continuous stirring to ensure even distribution. The resulting mixture was poured into an aluminum mold and left to solidify at room temperature for 5 to 10 min. Table 2 shows the list of biocomposites PCL samples (Fig. 6b) with varying MCC concentrations (0.2, 0.5, and 1.0 wt%). The “untreated” sample incorporated into the PCL refers to the raw materials, specifically the ground durian rind before any treatment has been applied.

Fig. 6

(a) molten PCL (b) PCL-based biocomposite

Table 2 List biocomposite samples with different MCC concentrations

Mechanical Properties Characterization

The mechanical properties of the PCL/MCC biocomposite were evaluated using a Shimadzu universal testing machine (UTM) in accordance with the ASTM D638 [33] standard procedure with some modifications. The PCL/MCC biocomposite samples were mounted in the UTM, and the tester’s grips were positioned at a specified grip separation of 102 mm. The PCL/MCC biocomposite samples were then subjected to tension until failure at a constant speed of 0.5 mm/min [34]. The fractured surface of the biocomposite was examined under a FEI, Quanta 400 FESEM. The samples were coated with platinum to ensure proper electrical conductivity, as reported previously [35].

Biodegradability Test

To examine the biodegradability of the biocomposites, samples were cut to 3 cm × 2 cm, with an average weight of 2 g (Fig. 7a). The initial weight of each sample was measured and recorded. The samples were then buried in the soil at a depth of 2 cm, and the pH was measured to be 6.67 [36]. The samples were left to degrade under aerobic conditions 7, 14, and 28 days. The moisture of the soil was maintained with distilled water (Fig. 7b) [37]. At every specified interval duration, the samples were retrieved from the soil, cleaned with distilled water, dried, and weighed. The biodegradability of the samples was evaluated using Eq. (5).

$$\:\text{B}\text{i}\text{o}\text{d}\text{e}\text{g}\text{r}\text{a}\text{d}\text{a}\text{b}\text{i}\text{l}\text{i}\text{t}\text{y}\:\left(\text{\%}\right)=\frac{{W}_{1}-{W}_{2}}{{W}_{1}}\times\:100\text{\%}$$

(5)

Fig. 7

Biocomposite samples (a) prepared for biodegradability test, and (b) buried in moist soil maintained with distilled water

where $\:{W}_{1}\:$ and $\:{W}_{2}$are the weight of sample before and after biodegradability test.

Thermogravimetric Gravimetric Analysis (TGA)

The thermal stability of the biocomposites was determined using a STA 8000 Simultaneous Thermal Analyzer (Perkin Elmer, Waltham, MA, USA). STA measures the change in weight of a sample as a function of temperature and the data will be used to calculate the maximum mass decomposition corresponding to temperature. The dried composite sample weighing between 7 and 17 mg was placed in a sample pan and heated in an environment of nitrogen gas to ensure the change in weight was due to thermal degradation. The heating rate was 20 °C/min, and the temperature range was 50 to 600 °C [38].

Result and Discussion

Durian Husk Moisture Content

The moisture content was measured on the durian husk (DH) and the white sac (WS). Table 3 presents the total weight loss for both DH and WS samples, revealing an average moisture content of 85.19 ± 0.19% for WS, marginally higher than that of the DH, which was recorded as 83.65 ± 0.54%. The results are quantitatively similar to those reported by Pusparizkita et al. [39], who obtained an average moisture content from DH dried in an oven at 100 ℃, with the highest moisture content of 84.15% for the WS and 82.98% for the outer skin. In this study, WS also has a higher moisture content and lower dry matter content than DH [40]. Conversely, DH has a lower moisture content because it contains more dry matter, such as protein, starch, and cell wall material obtained from photosynthesis and movement in the phloem [41].

Table 3 Moisture content of DH and WS

Effect of Bleaching Agent on Cellulose Yield

The cellulose extraction is initiated by utilizing the durian husk (DH) of the durian rind with a particle size smaller than 1 mm. The strong cohesion bond between lignin, hemicellulose, and pectin, resulting from the β-1,4 glycoside bond and amorphous structure, was disrupted by a 2% (w/v) NaOH solution for 3 h during the alkaline treatment [42]. The product of the alkaline treatment, including cellulose compounds, forms a brown-colored precipitate attributed to the presence of remaining lignin.

Following the alkaline treatment, the bleaching process was carried out, making the alkaline-treated sample brighter. According to Putri and Kurniyati [43], the bleaching process aims to eliminate the remaining lignin from the cellulose. Color-absorbing molecules, including chromophores, undergo oxidation during bleaching, making them water-soluble and polar. Consequently, the extracted cellulose achieves a whiter and brighter color through the bleaching process.

The results of durian husk powder extraction, as depicted Table 4, demonstrate that the yield of extracted cellulose from DH using sodium chloride (NaCl) and glacial acetic acid was 30.36%. However, using sodium hypochlorite (NaClO) as the bleaching agent resulted in a lower cellulose yield, specifically, 23.82%. These results align with the research conducted by Xing et al. [44], who used the same raw materials but different alkaline and bleaching solutions, resulting in a yield of 32.23%, and the research conducted by Penjumras et al. [45], which utilized durian husk and a different treatment method to obtain a cellulose yield of 33.12 ± 0.108%. It is apparent that NaClO may be more potent in eliminating non-cellulosic compounds from cellulose.

Additionally, using NaClO as a bleaching agent increases the brightness of the cellulose (Fig. 4b), turning the sample white. However, this effect does not occur when using NaCl and glacial acetic acid. The primary reason is that NaClO is a potent bleaching and oxidizing agent [46]. During the bleaching treatment with NaClO, the lignin network is partly degraded into soluble compounds and partly separated from cellulose by breaking the lignin-carbohydrate bonds [47]. It efficiently degrades color-causing substances [48], such as lignin, demonstrating excellent non-lignocellulosic removal efficiency in bleaching treatment. Therefore, more non-cellulosic substances in the raw material led to an enhancement in carboxyl and carbonyl content, as well as an improved whiteness index [49]. On the other hand, sodium chloride is a salt, not an oxidizing agent, and requires the addition of acetic acid to fully penetrate the durian husk [50]. During bleaching, acetic acid is added to sodium chloride to alter the pH to 3 to 4, creating a bleaching environment needed for lignin removal. Since acetic acid is weak, and the amount added to NaCl is minimal, which explains why NaClO is more effective than the NaCl/acetic acid mixture.

In summary, the effectiveness of NaClO as a bleaching agent was compared to mixture NaCl/acetic acid mixture in terms of cellulose yield and purity. The results showed that the NaCl/acetic acid mixture produced a higher cellulose yield of 30.36%, outperforming NaClO, which yielded 23.82%. However, it is crucial to consider additional factors, such as cellulose purity and experimental conditions, when determining the better bleaching agent. It is evident that NaClO is effective in producing high purity cellulose, as indicated by the bleached durian husk turning white due to the elimination of color-causing substances. Although NaClO is a powerful and effective bleaching agent, the specific conditions under which it was used in this study might have led to a lower cellulose yield than the NaCl/acetic acid mixture due to its aggressive action. However, the bleaching effect was better with NaClO than the NaCl/acetic acid mixture.

Table 4 Calculated cellulose yield based on different type of bleaching agent

Effect of Acid Concentration on MCC Yield

One of the crucial factors in the extraction of MCC during acid hydrolysis is acid concentration. The MCC yield significantly decreased when the oxalic acid concentration increased from 30% (w/v) to 65% (w/v), as evident in Table 5. The highest MCC yield of 47.35% was obtained at 30% (w/v), while 65% (w/v) of oxalic acid resulted in a lower MCC extraction yield of 34.39%. These findings confirm that a high oxalic acid concentration is unfavorable for MCC extraction because it leads to the complete degradation of cellulose, producing sugar molecules [51]. According to Tang et al. [52], a higher acid concentration can cause a severe breakdown of cellulose into sugars. Conversely, a lower acid concentration can result in poorly aggregated and scattered fibers. Different raw materials require relatively narrow and specific acid concentration ranges to achieve optimum results. Additional parameters, such as cellulose-to-acid ratio and extraction from various parts of the durian husk, were also considered as they may affect the MCC yield. Further research could be done to optimize the MCC yield by reducing the acid concentration to less than 30% (w/v), as lower acid concentrations have resulted in higher MCC, as shown in Table 5.

Table 5 Calculated MCC yield with different acid hydrolysis parameters

Effect of Cellulose-to-Acid Ratio on MCC Yield

Two cellulose-to-acid ratios (1:10 and 1:50) were studied, showing MCC yields ranging from 34.39 to 72.59% for both DH and WS samples (Table 5). When the cellulose-to-acid ratio decreased from 1:50 to 1:10, the MCC yield for DH increased from 47.35 to 72.59%. This result is in line with a study where MCC extracted from sweet sorghum using hydrochloric acid with a cellulose-to-acid ratio ranging from 1:10 to 1:30. Notably, the study reported that a lower cellulose-to-acid ratio appeared to favor a higher MCC yield [53]. This implies that a higher cellulose-to-acid ratio facilitates the hydrolysis reaction more easily. However, the movement of more H⁺ ions towards the β-1,4-glycosidic link weakens the influence of the acid ion on the breakdown of glycosidic bonds. A similar finding was reported by Doan and Chiang [54], that the yield decreased dramatically from 52 to 31% when the cellulose-to-acid ratio increased from 1:30 to 1:50 (mL/g). This occurs because a higher cellulose-to-acid ratio could enhance the degree of cellulose hydrolysis.

MCC Yield from DH and WS

This section discusses the yield of MCC extracted from different parts of the durian rind, i.e., the durian husk (DH) and the white sac (WS). It is apparent that each part yields a different amount of MCC. Based on the results shown in Table 5, with the same cellulose-to-acid ratio and acid concentration, the yield of MCC extracted from DH-30B (72.59%) is higher than that of WS-30B (66.10%). These yield values are lower than that reported by Baruah et al. [55] of 83%, where the study used sulfuric acid. However, the results obtained in the present work are acceptable because, in most of the investigations, the MCC yield does not affect the crystallinity index [56].

The higher yield of MCC extracted from DH indicates that its cellulose has more crystalline regions compared to WS. DH experiences faster hydrolysis during acid hydrolysis due to its higher dry content, as mentioned in Section III.A and non-cellulosic compounds, producing higher crystalline regions. Moreover, extracting MCC from cellulosic materials through acid hydrolysis breaks down the β-1,4 glycosidic linkages that bind the glucose monomers in cellulose and cause the cellulose molecules to hydrolyze [57]. Hydronium ions infiltrate the cellulosic material during hydrolysis, primarily disrupting the non-crystalline areas around the microfibrils and those embedded between them. By targeting the amorphous domains, the hydrolysis process can achieve faster reaction kinetics, resulting in a swift reduction of the degree of polymerization while maintaining the crystalline area untouched. When the well-structured crystalline regions achieve the maximal hydrolysis yield, further depolymerization will impede because the hydronium ions find it challenging to reach the crystalline region [58].

In summary, the cellulose yield and MCC yield were affected by various factors. Figure 8 summarizes the cellulose and MCC yield based on different conditions and parameters during the extraction process. The yield of cellulose was notably affected by the bleaching agent used. Sodium hypochlorite (NaClO) produces lower cellulose yield than sodium chloride and acetic acid. Additionally, the acid concentration and cellulose-to-acid ratio were major factors affecting the MCC yield during acid hydrolysis. Figure 8(b) and (c) demonstrated that a lower acid concentration of 30% (w/v) and a lower cellulose-to-acid ratio of 1:10 resulted in a much higher MCC yield of 72.59%. Furthermore, the results show that the yield of MCC extracted from WS is lower than that from the DH under similar acid concentrations and cellulose-to-acid ratio.

Fig. 8

(a) Cellulose yield based on different types of bleaching agents (b) MCC yield with different acid concentrations (c) MCC yield with different cellulose-to-acid ratios (d) MCC yield from different parts of durian rind with same acid concentrations and cellulose-to-acid ratio

Therefore, the optimal parameters and conditions for extracting MCC from DH and WS at higher yields are an acid concentration of 30% (w/v), a cellulose-to-acid ratio of 1:10, and the use of NaClO as a bleaching agent. However, it is important to note that the ideal parameters for MCC extraction may differ based on the raw material utilized and the specific extraction process conditions.

Functional Group in Extracted MCC

Figure 9 represents the FTIR spectra obtained for both untreated and MCC samples extracted from DH and WS using different concentrations of oxalic acid [30, 50, and 65% (w/v)]. The spectra revealed peaks at 3340 cm⁻¹ and 3441 cm⁻¹, indicating the presence of -OH groups, the predominant functional groups in cellulose [59; 60]. The intensity of these peaks deepened after acid hydrolysis and pretreatment, suggesting an increase in the hydrophilic characteristics of cellulose, primarily due to a higher concentration of hydroxyl groups in the fibers, as reported by Fitriani et al. [26].

Fig. 9

FTIR spectra of (a) untreated DH sample and MCC extracted from durian husk (DH) using 30%, 50% and 65% (w/v) oxalic acid, and (b) untreated DH and WS samples and MCC extracted from DH and WS using 30% (w/v) oxalic acid

The absorption bands observed at 2990 cm⁻¹ to 2882 cm⁻¹ represent the symmetric and asymmetric vibrations of C-H bonds [61]. The sharper peaks in the MCC samples indicate enhanced exposure of the cellulose components. The band corresponding to hemicellulose is visible between 1765 cm⁻¹ and 1715 cm⁻¹, attributed to the C = O stretching of carbonyl (ketone) groups [62]. The absence of peaks in the MCC samples, except for the DH-30B and WS-30B at 1734 cm⁻¹ (Fig. 9b), can be attributed to the removal of hemicellulose through pretreatment. The sharp peak at 1734 cm⁻¹ in DH-30B represents acetyl and uronic ester groups in hemicellulose or the ester linkage of the carboxylic group in ferulic and p-coumaric in lignin and/or C = O group in hemicellulose [63]. This indicates that the complex structure of the biomass can affect the effectiveness of hemicellulose removal even under standardized conditions and parameters during alkaline treatment.

Additional bands detected in the FTIR spectrum, specifically at 1632 cm⁻¹ for all samples (untreated and MCC), resulted from H-O-H bonding as the fibers absorb water. Despite undergoing a suitable drying procedure, removing water from the fibers proved challenging [6; 64]. Recent studies reported by Akhavan-Kharazian and Izadi-Vasafi [65] on the nanocellulose also proved the occurrence of hydrogen bonding in cellulose. The peaks observed in the FTIR spectrum, ranging from 1405 cm⁻¹ to 895 cm⁻¹ are indicative of cellulose presence and were detected in MCC samples but not significantly in untreated samples. Previous studies reported by Fazeli et al. [66] have shown that the peak indicates the presence of the CH₂ group in cellulose at 1405 cm⁻¹. In all MCC samples, peaks were observed at 1368 cm⁻¹ due to the twisting vibration of C = O and C-H bonds [67], and the band at around 1313 cm⁻¹ represents the CH₂ wagging vibration within cellulose [45].

Furthermore, the visible peak at approximately 1157 cm⁻¹ signifies the presence of glycosidic bonds within the cellulose chain. The absorption peak between 1027 cm⁻¹ represents the vibration of the C-O-C bond, a bond found in cellulose [6]. The peak at 895 cm⁻¹ signifies the C-O-C bonds within the β-1,4-glycosidic linkages of cellulose, as noted by Kusmono et al. [68].

The narrow absorption range between 1423 cm⁻¹ to 890 cm⁻¹ demonstrates that the cellulose structure of MCC samples comprises a more ordered structure than a disordered structure. This finding provides evidence that a high yield of MCC is being generated [69].

In summary, the FTIR results suggest that the cellulose structure remained intact after alkaline and bleaching treatment. The presence of hemicellulose peaks in the treated samples indicate that removing hemicellulose is challenging due to its complex structure. The absence of lignin peaks confirms the successful removal of lignin. Overall, the FTIR peaks support the successful cellulose extraction and suggest that further research should be conducted on the alkaline treatment of durian husk to explore hemicellulose removal.

Crystallinity Index of Extracted MCC

X-ray diffraction was employed to investigate the crystalline properties of cellulose treated with different acid concentrations and bleaching agents. The comparison of the XRD spectra for treated cellulose is presented in Fig. 10. The sharp and broad peak ranging from 22° to 24° suggests that all the extracted MCC exhibit a high degree of crystallinity. These results are similar to the work reported by Sainorudin et al. [70] on the MCC extracted from coconut coir, banana stem, sugarcane bagasse, and pineapple leaves. In the present work, DH-30B and WS-30B, which were treated with sodium hypochlorite, exhibited three peaks around 2θ = 16°, 22°, and 34°, corresponding to the (110), (200), and (004) diffraction planes. These peaks represent the typical crystalline structure of cellulose I [71]. Moreover, DH-30 A, DH-50 A, and DH-65 A, which were treated with sodium chloride and acetic acid, displayed three well-defined peaks around 2θ = 12°, 20° and 22°, corresponding to the planes of (1$\:\stackrel{-}{1}$0), (110) and (020). These three peaks are attributed to the antiparallel structure of cellulose II [72]. It is noteworthy that cellulose I and cellulose II are two different types of cellulose. The differences in XRD patterns between those samples indicate variances in cellulose structure due to the different bleaching agents employed.

Fig. 10

X-ray diffraction (XRD) spectra of DH-30B, WS-30B, DH-30 A, DH-50 A, and DH-65 A

Table 6 tabulates the calculated crystallinity (Crl) using Eq. (4) with the intensity of cellulose I and cellulose II at 2θ between 21° and 22°, and the intensity of amorphous region (I_am) at 2θ between 16° to 18°. The crystallinity for all MCC samples ranges from 71.43 to 78.30%. In the batch treated with sodium hypochlorite, DH-30B has a higher crystallinity index of 77.54% compared to WS-30B (73.81%). This may be due to the higher content of dry matter such as protein, starch, hemicellulose, lignin, and cell wall material in DH-30B (Section III.A), which gives a higher rate of hydrolysis than in WS-30B. The hydrolysis process increases the crystallinity index of DH-30B by breaking down the amorphous region, leaving behind a higher proportion of crystalline parts [73].

The calculated crystallinity index is qualitatively similar to Nasution and Sitompul [74], in which MCC was extracted from the fiber of empty fruit bunch palm oil and gave a crystallinity index of 73%. Variations in cellulose characteristics derived from different sources can lead to differences in the crystallinity index. The parameters used for the acid hydrolysis process can also influence the MCC properties.

For the batch treated with sodium chloride and glacial acetic acid (DH-30 A, DH-50 A, and DH-65 A), the crystallinity decreased from 78.3 to 71.43% as the acid concentration increased from 30 to 65% (w/v). This is because cellulose consists of both amorphous and crystalline components, and the amorphous region is more susceptible to hydrolysis [75]. The amorphous region of cellulose can be easily broken down by a lower acid concentration of 30% (w/v) without further degradation of sugar, resulting in smaller-sized cellulose with a greater crystallinity index. However, the crystalline region may suffer damage when a higher acid concentration, such as 65% (w/v), is employed, leading to a decrease in the crystallinity index [76].

Based on the X-ray diffraction analysis of MCC treated with different bleaching agents and acid concentrations, the study showed the various peaks in the XRD spectra indicated different cellulose structures, with NaClO-treated samples producing cellulose I, while NaOH and acetic acid-treated samples exhibiting cellulose II. On the other hand, higher acid concentration resulted in lower crystallinity due to damage to the crystalline region, although the overall crystallinity differences were not significant.

Table 6 Crystallinity index of the extracted MCC

Morphology and Particle Size of Extracted MCC

Figure 11 illustrates the UV-Vis spectra of MCC in distilled water. In general, the transmittance of DH-30B (Fig. 11a) is higher compared to that of WS-30B (Fig. 11b). DH-30B achieved the highest transmittance value of 90% at around 300 nm, which could be attributed to the variations in particle size between DH-30B and WS-30B (Fig. 12e and f). The variations in transmittance due to differences in particle size were supported by previous research. Zhai et al. [77] found that the large-sized particle original fiber has lower transmittance of 92.8% compared to the homogeneous fiber suspension with a transmittance of 96.7%. Figure 11 shows that there is a slight decline at approximately 250 nm to 300 nm wavelength and this could be translated to increased interference in the UV range. Previous research on polymer/CNWs composite films reported that the agglomerations of CNWs might result in interference in the lower spectrum, which reduces the UV light transmittance [78]. Hence, areas with high MCC density in distilled water may resemble the agglomerated areas in polymer/CNWs composite films, producing an interference effect if the MCC and distilled water do not mix homogeneously. Furthermore, another study has suggested that nanoscale fillers with less than 25 nm sizes may prevent scattering and maintain the material’s transparency, resulting in high transmittance [79]. Since this research focuses on micro-sized crystalline cellulose, the slight decrease shown in the 250 to 300 nm wavelength region could be explained by bigger particle sizes (> 25 nm), which is consistent with the results that show the influence of particle size on transmittance. According to Handoko and Yusuf [79], the transmittance is influenced by the dispersion of the particle, including the presence of agglomerations. The percentage of transmittance is wavelength-dependent, and more light scattering occurs when the wavelength of light is either similar to or smaller than the diameter of the particle. Therefore, in the present work, it can be speculated that the lower transmittance of WS-30B indicates the presence of agglomerations in the extracted MCC, thus forming a barrier to the passage of light. Cellulose in crystal form has a propensity to interact with each other due to their increased surface area of reaction, leading to a greater likelihood of aggregation [80].

Fig. 11

Ultraviolet-visible (UV-vis) spectra of (a) DH-30B and (b) WS-30B

Fig. 12

Scanning electron micrographs of (a) DH-30 A at 300X magnification (b) DH-30 A at 50,000X magnification (c) DH-50 A (d) DH-65 A (e) DH-30B and (f) WS-30B

The average particle sizes and size distribution obtained through laser scattering for all MCC samples are listed in Table 7. It is crucial to note that the size distribution does not indicate the width or length of the particles; instead, it provides an apparent “size” for all particles, including individual particles and agglomerated or aggregated particles [81]. Compared to DH-50 A and DH-65 A, DH-30 A has the smallest average diameter, measuring 6.237 μm with a size distribution of 0.4178 μm. This finding parallels to El-Sakhawy and Hassan [82], in which the MCC is extracted from different agricultural residues with the average particle size ranging between 4.06 and 5.49 μm.

Table 7 Particle size distribution of the extracted MCC

Meanwhile, DH-50 A and DH-65 A have average diameters of 25.38 μm and 19.04 μm, with size distributions of 0.002334 μm and 0.2208 μm, respectively. This is associated with an increase in oxalic acid concentration from 30% (w/v) to 65% (w/v), which resulted in a suspension with increasing particle size [83], indicating that more amorphous regions are being removed with lower acid concentration during acid hydrolysis. These results confirm that oxalic acid concentration is one of the most important factors dictating the reaction. Other researcher [84], also found that smaller particle sizes could be produced with lower acid concentrations. Sriruangrungkamol et al. [85] extracted nanocellulose from cellulose through varying acid concentrations and reported that the smallest particle size (31 ± 12 nm) was obtained with 5% (w/v) oxalic acid concentration, while 10% oxalic acid produced the largest nanocellulose with a particle size of 91 ± 33 nm. Therefore, in the present study, an oxalic acid concentration above 30% (w/v) favors acid hydrolysis. In general, excessive acid concentration at 65% (w/v) can damage the crystal structure of fibers and may even cause the fibers to carbonize, resulting in smaller particle sizes. However, since oxalic acid is a weak acid, it requires a higher temperature during hydrolysis [24] to produce smaller particle sizes efficiently. In contrast, acid hydrolysis using sulfuric acid typically operates at lower temperatures. Therefore, a low concentration of 30% (w/v) might be suitable for small MCC particle extraction with a temperature of 90 ℃. In contrast, higher oxalic acid concentrations, such as 50% (w/v) and 65% (w/v), may require a higher temperature to produce smaller MCC particles.

Moreover, the significant variance in individual particle diameter and size distribution for the extracted MCC is due to larger agglomerated and aggregated particles coexisting with individual microcellulose particles in the extracted MCC. This discrepancy suggests that different particle sizes due to aggregation and agglomeration cause inconsistencies in particle size.

In comparison between the MCC particle size extracted from different parts of the durian husk using the same acid concentration of 30% (w/v), DH-30B has a smaller average particle size (8.215 μm) than that of WS-30B (15.10 μm). This is in line with the UV-vis spectra (Fig. 11) that indicate a higher transmittance in DH-30B than in WS-30B, suggesting that UV-vis can be used to correlate the particle size of MCC through the turbidity of the suspension. The small particle size in DH-30B explains that the MCC extracted from different parts of the durian husk (DH and WS) affects the particle size, shape, and agglomeration characteristics. Bhardwaj et al. [86] stated that the removal of non-cellulosic compounds such as lignin, hemicellulose, and other impurities through pretreatment results in substantial morphological alterations in their size dimensions and structure.

Relating the results from particle size analysis to the previous section (Section III.A), DH-30B has more impurities than WS-30B, including protein, starch, and cell wall materials. Hence, more impurities are being removed from DH-30B through pretreatment, significantly reducing its particle size.

MCC extracted through different pretreatment and acid hydrolysis conditions were examined under FESEM, revealing that the extracted MCC has an irregular and micro-sized rod shape with a slightly rough surface (Fig. 12). This is attributed to the breakdown of cellulose fibers during the acid hydrolysis process. The uneven surface of the extracted MCC results from the cleavage of cellulose’s amorphous areas, breaking the fibers into shorter lengths and producing considerably weaker interactions that form cracks on the surface [87].

According to Kian et al. [88], MCC exhibits a diameter ranging from 40 to 300 μm. Particle size distributions measured through laser scattering (Table 7) proved that the extracted MCC are in the range of micro-sizes. The diameter of the MCC was further analyzed using ImageJ software to confirm the successful extraction of micro-sized cellulose from DH and WS. DH-30 A sample has a particle diameter ranging from 8.117 μm to 50.635 μm (Fig. 12a); however, the average particle size in Table 7 is approximately 6.237 μm. The SEM image of DH-30 A under 300X magnification (Fig. 12a) primarily shows bigger particles, while under the magnification of 50,000X (Fig. 12b), it reveals numerous smaller particles ranging from 0.164 μm to 0.324 μm. These explain why the average particle size of 6.237 ± 2.975E-2 μm measured using laser scattering is not within the range between 8.117 μm and 50.635 μm, which was measured using image J. Smaller particles were not observed in Fig. 12a as they may hide behind the bigger particles. The particle diameter of DH-50 A (Fig. 12c) also exhibits in micro sizes, ranging from 7.734 μm to 31.722 μm. However, DH-65 A (Fig. 12d) differs from those observed in DH-30 A (Fig. 12a and b) and DH-50 A (Fig. 12c), exhibiting an irregular, clear, smooth, and nonporous surface and a smaller particle size range between 6.425 μm and 28.619 μm. This could be attributed to the non-cellulosic components around the fibers, such as hemicellulose and lignin, that dissolved in the chemical during the acid hydrolysis [89]. A similar result was reported by Zhao et al. [90], in which MCC was extracted from tea waste, and the results show that the MCC detected consisted of individualized, long smooth fibers due to the acid penetration into the amorphous areas during hydrolysis leading to the cleavage of the β-1,4-glucopyranose linkage within the cellulose.

For samples used to study the effect of MCC extracted from different parts of the durian husk, DH-30B (Fig. 12e) and WS-30B (Fig. 12f) appear as aggregated cellulose micro balls, with some particles having larger sizes and irregular, rough surfaces. This contrast with the micro-sized rod-like structure observed in DH-30 A, DH-50 A, and DH-65 A (Fig. 12a-d) could be attributed to the presence of non-cellulosic compounds, such as hemicellulose, identified through FTIR analysis in Section III.F. Tan et al. [91] and Adawiyah and Suryanti [92] similarly noted that MCC extracted from lembang (Grewia florida) and cotton waste via acid hydrolysis having long threads-like structure, rod-like particle, in contrast to the micro-ball morphology, and uniform particle size. DH-30B (Fig. 12e) appears as fragments of microfibers with particle sizes between 5.201 μm and 24.849 μm, while WS-30B (Fig. 12f) exhibits larger particle size, ranging between 7.892 μm and 24.957 μm. Notably, this matches the findings from UV-vis (Fig. 11) that the DH sample shows higher transmittance in suspension, speculating a smaller particle size.

Mechanical Properties of Biocomposites

MCC has been identified as an effective reinforcing agent that enhances the mechanical properties of the polymer matrix. Figure 13 reveals that the tensile strength changes when untreated durian husk powder and MCC are added to polycaprolactone (PCL). The tensile strengths increased when 0.2 wt% of untreated durian husk powder (DH/untreated and WS/untreated) and MCC extracted from DH (DH/MCC) was added to PCL (Fig. 13a). The tensile strength of WS/untreated further increases when the concentration is raised to 0.5 wt% (Fig. 13b) and begins to decrease at a concentration of 1.0 wt% (Fig. 13c). However, the tensile strength of DH/untreated and DH/MCC decreases at a concentration of 0.5 wt% and increases at a concentration of 1.0 wt%.

Fig. 13

Average tensile strength with different reinforcement weight fraction for (a) DH/untreated (b) DH/MCC (c) WS/untreated (d) WS/MCC

In contrast, WS/MCC shows a different trend than the other samples. WS/MCC0.2 (with 0.2 wt% MCC) has lower tensile strength compared to pure PCL, and it starts to decrease as the concentration of MCC increases (Fig. 13d). The biocomposite fabrication technique is the main cause of the decrease in tensile strength. According to Tharazi et al. [93], heating time and pressure affect the tensile strength of the biocomposite. Tensile strength increases when the heating time increases from 5 to 8 min. However, tensile strength decreases when the heating time exceeds 8 min. Additionally, high pressure can lead to low tensile strength because the pressure can potentially harm the matrix and result in inadequate matrix infiltration. As a result, in the present work, the decrease in tensile strength of WS/MCC reinforced biocomposite could be the inconsistency in controlling the heating time and pressure during the fabrication. PCL was melted in a water bath at 70 ℃ throughout the fabrication of the biocomposite process. The heating time required to melt the PCL was not precisely measured, resulting in some samples being heated longer than others. In addition, the PCL was manually pressed after being filled into the mold. As a result, several disturbances were introduced to the fabricated biocomposites. Additionally, the cooling rate is another factor that impacts the mechanical properties of the biocomposites. Hence, future work should carefully manage the pressure, cooling and heating time during the biocomposite fabrication to ensure that fibers or MCC are distributed evenly in the matrix.

Overall, MCC extracted from DH has the potential to enhance the tensile strength of the PCL-based biocomposite, whereas MCC extracted from WS does not exhibit the same improvement. When 1.0 wt% of DH/MCC was added to the biocomposite, the tensile strength improved from 1.923 MPa to 2.339 MPa, representing a 22% increase. This finding aligns with the results reported by Van Hai et al. [94], indicating that crystalline cellulose can indeed enhance the tensile strength of the biocomposite. However, it is crucial to note that a large amount of crystalline cellulose can impact the interaction between the polymer matrix and reinforcing agent due to inadequate particle dispersion and the formation of large agglomerates. Consequently, the tensile strength decreases at high concentrations of MCC. Although DH/MCC and WS/MCC biocomposites were incorporated with microcrystalline cellulose, WS/MCC has a lower tensile strength. This could be due to the higher crystallinity of DH/MCC biocomposite compared to that of the WS/MCC. Faisal et al. [95] stated that the mechanical properties of the biocomposite are impacted by the crystallinity of the reinforcing agent, with higher crystallinity improving the mechanical properties. Other than that, the mechanical properties are also affected by the larger particle size caused by the agglomeration of the hydroxyl group on the WS surface, as indicated by UV-vis and FESEM analysis in Section III.H. Similar findings were reported by Luddee et al. [96], who found that the tensile strength and elongation of the polylactic acid (PLA) incorporated bacterial cellulose composite significantly decreased when the bacterial cellulose particle size increased from 90 to 180 μm. According to Lu et al. [97], the enhancement in the mechanical properties of the biocomposite could be attributed to the hydrogen bonding between the hydroxyl group of the filler and the polymer matrix molecules, strengthening the interfacial interaction. In the present work, WS/MCC exhibited a lower tensile strength because the hydroxyl groups on the WS surface attempt to form strong hydrogen bonds with adjacent hydroxyl groups in the WS chain instead of forming strong bonds with the PCL polymer. The strong hydrogen bonds on surface lead to entanglements and agglomerations of the nanocellulose, resulting in lower tensile strength.

The untreated durian rind also has a larger particle size since the non-cellulosic compounds are not being removed. Therefore, a noticeable gap between the matrix and fibers was created, which implies a weak interaction between the matrix and the fibers, as shown in Fig. 14a. According to Huang et al. [98], the mechanical properties of the matrix can be enhanced by adding fibers. The fibers bear the majority of the tensile load applied on a sample and bridge cracks in the matrix, thereby slowing the growth of cracks by dissipating energy close to the crack tip. Hence, the gap between the untreated durian rind and PCL indicates the poor compatibility of the fiber-matrix bonding, resulting in lower tensile strength and Young’s modulus as the tension load required to break the fiber-matrix bond is minimal.

Fig. 14

SEM image of the cross-section of (a,b) DH/untreated1.0, (c,d) WS/untreated1.0, (e,f) DH/MCC1.0 and (g,h) WS/MCC1.0

The effect of reinforcing MCC and untreated durian husk on Young’s Modulus of the biocomposite was further explored in this study, and the results are illustrated in Fig. 15. Young’s Modulus decreases when untreated durian husk powder and MCC are added to the PCL. Notably, DH/MCC0.2 exhibits the lowest Young’s Modulus at 237.652 MPa, but it increases to 357.784 MPa when the MCC concentration reaches 1.0 wt%. The result is in agreement with an earlier work reported by Rose Joseph et al. [99], Young’s Modulus tends to increase with the addition of reinforcing agents. Moreover, Paul et al. [100] also found similar results where the Young’s Modulus increased from 1556 MPa to 2517 MPa when MCC was added to PLA. The strong compatibility of natural filler with the polymer matrix, facilitated by hydrogen bonding interactions, enhances the efficient stress transfer from the PCL polymer matrix to MCC, resulting in improved mechanical properties. However, in the present work, the Young’s Modulus of DH/untreated, WS/untreated, and WS/MCC remains lower than that of the pure PCL. A low Young’s Modulus in these biocomposites suggests weak intermolecular interactions between the reinforcing agent (untreated powder and MCC) and PCL [101]. Additionally, the decrease in Young’s modulus is attributable to the aggregation of the reinforcing agent at higher concentrations, driven by the large specific area and high surface energy of the particles in the filler. Hence, a higher concentration of reinforcing agent is likely to cause agglomerations, which lowers and reduces the mechanical properties. Besides that, during the fabrication of biocomposites, a concurrent and reversible agglomeration process occurs alongside the aggregation [102]. Furthermore, the crystallinity index of the reinforcing agent (Section III.G) also impacts the mechanical properties. The more ordered structure of DH/MCC compared to the untreated fibers and WS/MCC likely contributes to the superior mechanical properties of the DH/MCC, hence the biocomposites [103].

Fig. 15

Average Young's Modulus with different reinforcement weight fraction for (a) DH/untreated (b) DH/MCC (c) WS/untreated (d) WS/MCC

Moreover, the influence of additional MCC and untreated durian husk on the elongation at break of the biocomposite is depicted in Fig. 16. It can be clearly seen that the percentage of elongation at break fluctuates as the concentration of MCC and untreated durian husk increases, with the minimum and maximum elongation at break recorded at 0.557% and 0.934%, respectively. All samples exhibit an increase at the 0.2 wt% concentration of untreated durian husk and MCC, except for WS/MCC, suggesting that incorporating MCC and untreated durian husk at 0.2 wt% enhances ductility. The improved distribution of MCC in the PCL matrix contributes to the increased ductility at the 0.2 wt% concentration. However, this does not necessarily imply greater mobility of the PCL chain; it simply results in increased ductility. Furthermore, the addition of 0.5 wt% MCC reduces the ductility of the biocomposite, likely caused by the restricted movement of the PCL chain due to the introduction of a larger quantity of MCC [68].

Fig. 16

Average percent of elongation with different reinforcement weight fraction for (a) DH/untreated (b) DH/MCC (c) WS/untreated (d) WS/MCC

When the concentration is increased further to 1.0 wt%, the elongation at break increases again, except for WS/MCC (Fig. 16d). Other researchers have also demonstrated a decrease in elongation at break with the increasing concentration of the natural reinforcing agent. Sultana et al. [104] found that the elongation at break decreased with the increasing content of nanocellulose and cellulose from 1 to 5 wt% in PVA biocomposites. Similar findings were also reported by Noorbakhsh-Soltani et al. [105], where the maximum values in elongation at break were achieved for 5 wt% natural filler. However, the elongation at break decreased further as the filler concentration increased to 25 wt%. The low elongation at break is attributed to the brittle and rigid behavior of crystalline cellulose. Mohan Bhasney et al. [11] also found that adding MCC to PLA reduced the tensile strength while increasing Young’s Modulus. In comparison to pure PLA, the tensile strength and percentage elongation decreased to about 32 MPa and 10.3%, respectively, upon incorporating 0.1 wt% MCC. When the MCC concentration increased to 0.5 wt%, the percentage elongation and tensile strength decreased by approximately 34% and 54%, respectively. Hence, in the present work, the elongation at break is also affected by the amount of reinforcement added, the dispersion of the reinforcement within the PCL, and the interaction between the matrix and the reinforcement [106]. MCC tends to agglomerate, resulting in poor dispersion and interaction in the PCL matrix. As a result, the elongation at break decreases.

Table 8 summarizes the tensile strength, Young’s Modulus, and percent of elongation for PCL-based composite. The MCC extracted from DH has shown promising potential as a natural reinforcing agent to improve the mechanical strength of biocomposite. This improvement is consistent with earlier research showing that crystalline cellulose, such as MCC, can significantly increase the tensile strength of polymer matrix. The present study also showed that the MCC extracted from DH tends to increase the Young’s Modulus compared to pure PCL biopolymer, indicating improved stress transfer within the PCL-based biocomposite. Though this work has shown enhancement of mechanical properties, further study is recommended to improve the mechanical properties of PCL-based composites. One promising approach involves modifying crystalline cellulose to enhance its interface compatibility with the polymer matrix. For instance, Jin et al. [107] found that the compatibility of crystalline cellulose and polylactic acid (PLA) can be enhanced through surface modification because the alkyl side chains grafted onto the crystalline cellulose surface and the PLA side chain intertwined and interpenetrated, increasing the intermolecular interaction force and the effective contact area between the modified crystalline cellulose and PLA. Huang et al. [108] also performed surface modification to increase the hydrophobicity and dispersibility of crystalline cellulose using stearic acid. The modified crystalline cellulose was incorporated into PLA film-forming at different concentrations. The modified crystalline cellulose forms a stable network with PLA by entanglement and co-crystallization of the molecular chains. This interaction strengthened the bonding between the reinforcing agent and polymer matrix, leading to higher elastic modulus and tensile strength.

Table 8 Calculated tensile strength, Young’s Modulus, and percent of elongation with standard deviation

Analysis of Fracture Surfaces of PCL/MCC Biocomposite

The SEM images of the fracture surfaces after tensile testing with fiber pullout, some voids, waviness, and fiber fracture for WDH/Untreated and MCC-reinforced PCL composites (MCC/WPO) are shown in Fig. 14. Fig. 14a-d show the PCL incorporated with untreated DH and WS, while Fig. 14e-h illustrated the PCL incorporated with MCC extracted from DH and WS. The presence of small void space can be observed in all samples. It is clear that fiber pullouts are not the main cause of this small void space. These void spaces, also known as gas porosities [109], are likely attributed to the shrinkage of the sample from molten to solid state during the biocomposite fabrication process. On the other hand, the void space within the biocomposite also indicates that the MCCs are not evenly distributed [110].

The SEM images depicted that the fracture parts have a rough surface due to the addition of untreated durian rind and MCC. The untreated fibers are dispersed throughout the PCL matrix in DH/untreated biocomposite (Fig. 14a). However, the MCC in DH/MCC1.0 cannot be observed due to the small micro-sized MCC. The PCL biocomposite incorporated with DH/untreated1.0 exhibited an agglomeration and large porosity on its surface (Fig. 14a). However, there are still noticeable gaps between the matrix and fibers (indicated by arrow), which implies a weak interaction between the matrix and the fibers [35].

Compared to DH/MCC1.0 and WS/MCC1.0, the mechanism of fiber pullout is more noticeable in DH/untreated1.0 and WS/untreated1.0. This could be attributed to fiber debonding during fracture, leading to an extended crack propagation length during tensile testing and, consequently, an enhancement in fracture toughness [111]. According to Pavan et al. [112], when the particle is pulled and slid over the PCL matrix, a tiny peak will be created at the location where the particle is ejected. Smaller particles like MCC are easily removed, whereas large particles like untreated durian husk can withstand the load and leave an imprint before being pulled out from the surface.

Biodegradability Test

Research on biodegradation characteristics is crucial for understanding their degradation process and promoting sustainability. This study conducted soil burial experiments for pure PCL and the PCL-based biocomposites containing 0.2 wt%, 0.5 wt%, and 1.0 wt% of MCC extracted from DH and WS. The percentage of weight loss for each biocomposite over biodegradation time is plotted in Fig. 17. After 7 days of incubation in soil, only pure PCL and WS/MCC (0.2, 0.5, and 1.0 wt%) experienced weight losses, ranging between 0.07 and 0.15%. By day 28, the results revealed that the weight losses of the WS/MCC biocomposites were all higher than that of pure PCL. Therefore, the presence of MCC in biocomposites increased the degradation of PCL as these biocomposites exhibit more significant weight loss than that of pure PCL, as evidenced by the steeper percentage weight loss curve in Fig. 17.

Fig. 17

Biodegradability test weight loss in PCL and PCL-based biocomposites with different reinforcement weight fraction

When comparing the weight loss of WS/MCC with DH/MCC biocomposites at 0.2, 0.5, and 1.0 wt.% MCC concentrations, WS/MCC demonstrates a higher weight loss percentage than DH/MCC. This is because cellulose, which is a molecule consisting of two regions, namely highly oriented molecules known as ‘crystalline cellulose’ and less oriented molecules called ‘amorphous cellulose’ [113], can significantly impact the degradation ability based on its physical and chemical properties, including morphology, orientation, and crystallinity. Research indicates that highly crystalline cellulose is more challenging to degrade. The XRD analysis shows that DH-30B has a higher crystallinity index than WS-30B (Table 6); hence, DH/MCC biocomposite exhibits higher crystallinity than WS/MCC biocomposite, making it more resistant to microorganism attacks [114]. Higher crystallinity in DH/MCC indicates that the structure of DH is highly ordered and tightly packed compared to WS, which has lower crystallinity, is less ordered, and is more accessible to microorganisms [115]. This explains the variations in weight loss between DH/MCC and WS/MCC.

However, the negative weight loss observed in the DH/MCC at 0.2, 0.5, and 1.0 wt% MCC concentrations indicates a weight increase upon incubation in soil. Durian husk, containing natural fibers, exhibits hydrophilic properties [114] and can easily absorb moisture [116]; resulting in the weight increase. Research suggests that moisture can enter biocomposites through three main mechanisms [94]. Firstly, water molecules diffuse into the tiny gaps between polymer chains. Secondly, capillary transport occurs through openings and imperfections resulting from inadequate wettability between the matrix and fibers. Thirdly, the transfer happens via microcracks that develop in the matrix due to fiber swelling or during the mixing process with the fiber [117]. Therefore, the hydrophilic properties of natural fibers can affect the weight loss performance of biocomposites.

In contrast, WS/MCC showed a positive weight loss, probably due to the degradation caused by microorganisms. In a wet environment, water molecules infiltrate through the microcracks caused by fiber swelling during water absorption, reducing the interfacial adhesion between the polymer matrix and fiber [118]. This facilitates the transport of water containing microorganisms and their enzymes within the polymer matrix [119; 120], thereby enhancing the biodegradability of the WS/MCC biocomposites.

As illustrated in Fig. 17, the biocomposites demonstrated a degradation rate of approximately 0.07% after 7 days and increased to 0.42% over 28 days. The considerably low degradation rate may be attributed to the hydrophilic properties of the MCC, which increased water uptake in the biocomposites, thereby preventing a reduction in weight loss. Additionally, due to time constraints, our study only covered a 28-day period, which may not have been sufficient for observing a more significant degradation of the biocomposite. Other studies [121] have conducted the biodegradability test over more extended period and observed that biocomposite degrades more quickly over longer times, such as 5 months. However, the initial degradation rate was very slow, consistent with our findings. The low degradation rate in our study could also be due to the type of soil used, which is natural soil. Other researchers have found that the weight of biocomposite decreases more rapidly in compost soil than in natural soil due to the high humidity of compost soil and neutral pH of 7, which is an optimal condition for microorganism growth [122].

In summary, the biodegradability results clearly indicate that PCL/MCC biocomposites do not incorporate any adverse ecological effects. In other words, PCL/MCC biocomposites are entirely biodegradable.

Thermal Stability

TGA was performed to examine the thermal properties of the PCL-based biocomposite and monitor changes in weight%. This technique is helpful in identifying the maximum temperature that scaffolds can withstand during thermal processing and their homogeneity. Figures 18 and 19 shows the thermogravimetric (TG) thermogram and first-order derivative (DTG) of the PCL-based biocomposites. The onset, maximum, and endset degradation temperatures, determined through TGA analysis, are presented in Table 9. Virgin PCL degrades in a single step, beginning at 410.12 °C and ending at 462.38 °C (Table 9). On the other hand, the addition of untreated durian rind and MCC to PCL increased the onset degradation temperature to a range between 417.77 and 420.74 °C. The lower onset degradation temperature of virgin PCL suggests that it reached the exothermic peak faster than the PCL-incorporated natural reinforcing agent [123]. Ye et al. [124] also observed similar results, where the thermal stability of the PCL significantly improved with the addition of MCC, suggesting that adding MCC prevents PCL from degrading and enhances its thermal stability. Both virgin PCL and biocomposites displayed a single degradation curve in the TGA curve, indicating the excellent miscibility of the biocomposites [125].

Fig. 18

Mass (%) versus temperature curves of (a) PCL/DH and (b) PCL/WS biocomposites obtained from TGA experiments

Fig. 19

DTG versus temperature curves of (a) PCL/DH and (b) PCL/WS biocomposites obtained from TGA experiments

In PCL/MCC biocomposites, the onset degradation temperature increased with higher MCC content. The thermal degradation temperature rises due to the increase in molecular weight caused by the cross-linking interaction between the PCL matrix and MCC or the molecular chain extension of the matrix itself [126]. Consistent with these findings, Abbas et al. [127] also noted that the enhanced thermal stability is attributed to the positive interactions between PCL and MCC. Hydrogen bonds formed between the hydroxyl groups of MCC and carbonyl groups of PCL prevent the fast dehydration of the MCC. Additionally, gases and char produced during cellulose breakdown interact with the solid PCL, further enhancing the thermal stability [128]. Thus, MCC serves as a stabilizer that accounts for the increased thermal stability. In contrast, the thermal stability of the biocomposites incorporated with untreated durian rind slightly decreases when the filler loading increases, mainly due to the interaction between the natural reinforcing agent and the PCL matrix. The inadequate interfacial adhesion between the hydrophobic PCL and hydrophilic fiber leads to poor compatibility between the two materials. As the concentration of untreated durian rind increases, the incompatibility increases, resulting in reduced thermal stability [129].

According to the TGA data, biocomposites with untreated durian rind have higher thermal stability than PCL/MCC biocomposites. The untreated durian rind consists of non-cellulosic compounds, such as hemicellulose (30.7%) and lignin (15.6%), which affect the thermal stability of the biocomposites. According to Gond et al. [130], treated fiber exhibits lower thermal stability than untreated fiber due to the elimination of protective hemicellulose and lignin, which serve as a thermal barrier. These findings demonstrate that the amount of lignin, hemicellulose, and cellulose in a biocomposite affects its initial decomposition temperature and thermal stability.

Another important thermal property of the biocomposite is the temperature corresponding to the maximum degradation rate of weight loss (T_max). T_max is identified as the peak found in the first derivative curve of the TGA thermograph. From Figs. 18 and 19; Table 9, it is evident that raising the untreated durian rind or MCC content in the PCL biocomposite increases the maximum degradation temperature, except for DH/untreated. The increased T_max indicates that thermal stability improved due to the incorporation of a reinforcing agent as a barrier layer. These layers restrict the release of generated degradation gasses and heat transmission, thereby improving the thermal stability of biocomposites [131]. However, the increase in T_max is insignificant, at around 2 to 3 °C only. Similarly, Onbattuvelli et al. [132] have found that degradation temperature is usually affected by increasing the reinforcing agent content in PCL-based composites. They noted that the degradation temperature might change with greater concentrations of crystalline cellulose, such as 11 wt%, mainly because the higher degradation temperature is attributed to the increased interfacial area between crystalline cellulose and polymer.

In summary, incorporating untreated durian rind and MCC as a reinforcing agent in PCL-based biocomposites improved the thermal stability of the biocomposites. In addition, increasing the reinforcing agent content increases the thermal stability of the biocomposites.

Table 9 Onset, maximum, and endset degradation temperatures of PCL-based biocomposites

Conclusion

This work focuses on extracting microcrystalline cellulose (MCC) from durian husk and explores its potential application as a reinforcing agent within biopolymer matrix. Extraction of MCC from whole durian husk (DH) and white sac (WS) through acid hydrolysis using different concentrations of oxalic acid and varying cellulose-to-acid ratios was investigated. The particle sizes of extracted MCC ranged from 6.237 μm to 25.38 μm. Findings from this work address the existing knowledge gap, demonstrating that lower oxalic acid concentrations at 30% (w/v) and lower cellulose-to-acid ratio (1:10) provide a better MCC yield. The highest extracted MCC yield was 72.59% for extraction from DH and 66.10% for extraction from WS. It is suggested that further reducing the acid concentration can enhance the yield, as some researchers found that oxalic acid concentrations lower than 30% (w/v) could provide a higher MCC yield.

Additionally, FTIR confirmed the presence of cellulose in the extracted samples. However, remnants of hemicellulose were found in samples. Incomplete removal is potentially due to the complex structure of hemicellulose, posing difficulty for its complete elimination. UV-vis analysis indicated that the transmittance of the MCC extracted from different parts of the durian husk varied, with the lower transmittance between 250 nm and 300 nm due to considerably larger particle size and the presence of agglomeration blocking the light. Furthermore, XRD indicated that the crystallinity index of the extracted MCC ranges between 71.3 and 78.3%.

PCL-based biocomposites with varying MCC concentrations were investigated. The tensile testing demonstrated that reinforcing MCC extracted from DH can improve the mechanical properties of PCL-based biocomposite, with the tensile strength improving by 12% and the Young’s Modulus improving by 15% when 1 wt% of MCC was added to the biocomposite. Since the tensile strength and elongation of the biocomposite demonstrated improvement with the addition of 1 wt% MCC, further work on the maximum concentration of MCC in the biocomposite without compromising the mechanical properties of the biocomposite can be explored. The cross-section of the biocomposite fracture surface indicated the contribution of the MCC during the pullout. Furthermore, the biodegradability test tracked the weight of the PCL/MCC biocomposites, revealing a weight loss ranging between 0.58 and 0.35% after 28 days.

Overall, based on the current study, the optimal conditions for MCC extraction are an acid concentration of 30% (w/v) and a cellulose-to-acid ratio of 1:10. These parameters offer the highest MCC yield. Nevertheless, it is worth exploring the extraction yield by further reducing the acid concentrations. When considering different parts of the durian husk, DH provides better results than WS because of its ability to provide higher MCC yield, better UV-vis transmission, smaller particle size, higher crystallinity, and improved mechanical properties when incorporated into PCL. Moreover, sodium hypochlorite (NaClO) served as a better bleaching agent compared to sodium chloride (NaCl) and glacial acetic acid (CH₃COOH) for extracting cellulose from durian husk as it gives a higher yield and brighter color cellulose. Finally, the thermal stability of PCL after incorporating untreated fiber and MCC was increased, and this was proved by the higher onset temperature obtained through TGA analysis.

Considering the fabricated biocomposite exhibits a brittle property, using a thicker mold made of stainless steel, which has lower malleability, is recommended. The aluminum mold used in this work is too thin and susceptible to scratching, resulting in an uneven surface for the fabricated biocomposite. Moreover, the biocomposite can be fabricated using a hot press machine, which can maintain a constant pressure applied to the biocomposite and melt PCL in a shorter duration than in a water bath. The optimal temperature, pressure, and pressing duration during the fabrication must be determined, as these parameters vary based on the chosen polymer matrix. Lastly, it is advisable to explore the impact on the mechanical properties of the biocomposite by incrementally increasing the MCC concentration in the PCL matrix. Modifications to crystalline cellulose are also recommended to investigate the interface compatibility with the polymer matrix and further impact on the mechanical properties of the biocomposites.