Extraction and Characterization of Micro-fibrillated Cellulose from Rice Husk Waste for Biomedical Purposes

This study aimed to report the extraction of micro-fibrillated cellulose (MFC) from rice husk (RH) through a series of processes including alkalization, bleaching, chemical hydrolysis, and mechanical treatment. The chemical structure, morphology, and crystallinity were assessed using Fourier Transform Infrared spectroscopy (FTIR), Scanning Electron Microscope (SEM), and X-ray diffraction (XRD). The results showed that alkalization was more effective in removing unwanted substances such as silica, hemicellulose, and lignin compared to bleaching. Chemical or mechanical treatment was more targeted towards removing the amorphous phase while fibrillating MFC. Further mechanical treatment significantly enhanced the crystallinity index (CI) of MFC, reaching 87.47%, while chemical treatment remained at 78.54%. The mechanical treatment led to a larger void size due to rigorous fibrillation, resulting in increased water retention during extraction compared to chemically treated MFC with a negatively charged surface. Crystal extraction through mechanical treatment disrupted the hydrogen bond, transforming cellulose crystal from triclinic  to monoclinic The comprehensive evaluation of MFC extracted from RH showed its potential for biomedical application.

Acid hydrolysis; Mechanical treatment; Micro-fibrillated cellulose; Rice husk

    Exploration of bio-based materials is a process for addressing global challenges such as the lack of petroleum, climate change, and unmanageable waste. Cellulose, a member of the polysaccharide family, is the most prevalent organic component generated from biomass. This material has diverse applications in industries, including textiles (Felgueiras et al., 2021), food (Suryanti et al., 2023; Mu et al., 2019), and filtration membranes (Rochardjo, Fatkhurrohman, and Yudhanto, 2021). Cellulose holds tremendous potential as a biomedical material due to its biocompatibility, biodegradability, water-retention capacity, renewability, and tunability, making it an ideal biopolymer (Hu et al., 2021; Iwamoto, Nakagaito, and Yano, 2007). Micro-fibrillated cellulose (MFC) offers advantages as a pharmaceutical excipient for extended and controlled drug release, as well as reinforcing hydrogels and composites in biomaterial implants (Pandey 2021).

The use of environmentally friendly and sustainable biomedical materials has become increasingly important in healthcare studies and development. Rice husk (RH) waste, the outer covering of rice grains resulting from the milling process, constitutes a natural fiber comprising cellulose, hemicellulose, lignin, and other compounds such as wax, pectin, and silica (Johar, Ahmad, and Dufresne, 2012). Studies showed that RH contained cellulose (25-35%), lignin (26-31%), hemicellulose (18-21%), and silica (15-17%) (Rezanezhad, Nazarnezhad, and Asadpour, 2013).

An extensive study has been conducted on the isolation of cellulose from biomass/natural fibers (Ng et al. 2015). This process aimed to remove lignin, hemicellulose, and other non-cellulose compounds that naturally encapsulate cellulose. Alkalization and bleaching are the most commonly used chemical treatments. These treatments not only remove non-cellulose components but also defibrillate cellulose fibers into micro-sized (Yuanita et al., 2015). Menawhile, mechanical treatment through grinding is a straightforward, energy-efficient process that requires simple equipment. Uetani and Yano (2011) explored the isolation of nano-fibrillated cellulose using a modified blender. In the mechanical grinding method, shear forces from the blades break hydrogen bonds, reducing cellulose cell wall size to the nanoscale (Abdul-Khalil et al., 2014). Therefore, this study aimed to compare the effect of chemical and mechanical treatments on the characteristics of micro/nano-fibrillated cellulose as a biomedical material. The resulting material from processed RH holds potential applications in various medical fields, including surgical paper, wound coverings, bone patch devices, and drug delivery systems.

2.1. Materials

      RH (Oryza sativa L.) was obtained from a local rice field in Wonogiri, Central Java, Indonesia. Sodium hydroxide (NaOH, 99.9%), Sodium chlorite (NaClO₂, 25%), and acetic acid (CH₃COOH, 99%) were procured from Merck, while sulfuric acid (H₂SO₄, 96.1%) was obtained from Mallinckrodt AR.

2.2. Treatment Preparation

In this study, virgin RH was subjected to sequential treatments by alkalization, bleaching, and hydrolysis or mechanical methods, to obtain the treated RH as RH_Al, RH_Al-Bl, and RH_Al-Bl-Ch or RH_Al-Bl-Mc, respectively. The following section provides the details for each process.

2.2.1. Pre-treatment

       MFC extraction from RH comprised 2 primary steps, namely pretreatment (washing, crushing, alkalization) and MFC isolation using either mechanical or chemical methods. Pre-treatment began with washing RH fibers in flowing water to remove visible debris, followed by soaking RH and drying it at room temperature for 24 hours. The soaked fibers were then crushed through a 40-mesh sieve and alkalinized in a 5 w/w% NaOH solution at a 1:25 w/v ratio for 2 hours while stirring at approximately 80°C. After draining and rinsing with distilled water, alkalized RH fibers were bleached for 2 hours at 70°C using 1.7% NaClO₂ in a buffer of 100 ml of 0.2 M acetic acid and 0.291 g NaOH. According to Iwamoto, Nakagaito, and Yano (2007), this bleaching process was repeated 5 times until the sample was cloudy white. Finally, the obtained product was rinsed with distilled water to remove residual lignin, hemicellulose, and chlorine ions. Observations were conducted at each step, including RH, fibers after alkaline treatment (RH_Al), as well as fibers after alkaline and bleaching treatments (RH_Al-Bl).

2.2.2. MFC Isolation

        Following the removal of lignin, hemicellulose, and other impurities, the pre-treatment was followed by isolating MFC through acid hydrolysis (RH_Al-Bl-Ch) and mechanical milling, (RH_Al-Bl-Mc). In this study, MFC results from both methods were compared to identify the most effective. Acid hydrolysis treatment was performed by adding 60% v/v sulfuric acid to the pre-treated RH fibers in distilled water (1/25 of w/v) and mixing at 45°C (Syafri et al. 2011). After 45 minutes, the hydrolysis was terminated by adding cold distilled water (about 15°C). For the mechanical milling treatment, water (98/2 of w/w) was added to the pre-treated RH fibers in a blender, and the blade was rotated between 11,000 - 12,000 rpm at room temperature for 20 minutes (Johar, Ahmad, and Dufresne, 2012).

2.2.3. Characterizations

        RH at each treatment stage was evaluated by using a Field Emission Scanning Electron Microscopes (FE-SEM) Quanta 650 EDAX EDS Analyzer at varying magnifications of 100 and 500 times. Crystallinity percentage was determined through hydrogen bonding analysis, using FTIR spectroscopy (Perkin Elmer 90325) and X-ray diffraction (XRD) with a Philips XRD. FTIR was also adopted to investigate the chemical structure, while XRD data was analyzed to assess polymorphism and the allomorph of the extracted MFC. The crystallinity ratio (CrR) and hydrogen bond energy (EH) for specific OH stretching bands in cellulose fibers were calculated by comparing absorbance peaks at 1372 cm^-1 (A1372) and 2900 cm^-1 (A2900) (Nelson and O’Connor 1964). Gaussian deconvolution peak separation methods were applied to XRD data to obtain the crystallinity index (CI) of cellulose fibers, calculated using the Segal Equation (Park et al., 2010).

3.1.  Physical Appearance

This study adopted alkalization, bleaching, hydrolysis, and mechanical treatment methods to extract MFC and separate it from unwanted impurities. These methods yielded distinct physical changes, particularly in color and texture.

Figure 1 Physical appearance of isolated husk rice fibers for (a) RH, (b) RH_Al, and (c) RH_Al-Bl

Figure 1 shows the significant morphological changes observed during MFC isolation process. A noticeable color shift was discovered in each phase, showing a reduction in impurities such as lignin, hemicellulose, and silica, hemicellulose, and silica. RH_Al sample maintained a dark color compared to RH, suggesting the presence of lingering lignin. In contrast, RH_Al-Bl sample was white, showing the effective removal of lignin during the bleaching steps.

3.2. Morphology Evolution

Figure 2a–e shows the SEM images at 100 and 500  magnification for each sample. MFC isolation process comprised unfolding from the sheet into separate fibrils, resulting in a reduction in fibril width and the formation of a network structure. In Figure 2a, SEM image of the as-received RH sample lemma surface showed papillae (Park et al., 2003) and hairs of various sizes (Chen, Xu, and Shivkumar, 2018). However, several areas that lack these structures, disintegrated during rice milling. The outer epidermis had an uneven linear furrow decorated with conical protrusions, each comprising plant cell walls with bundled cellulose and lignin. Silica, an impurity, was concentrated at the tip and shoulders of these protrusions (Chen, Xu, and Shivkumar, 2018). Figure 2b shows the retainment of ridged structures with furrows, indicating multiple layers of cell walls (Ludueña et al., 2011). Furthermore, the protrusions and hairs were removed, showing partial hemicellulose and lignin removal. Silica was visibly reduced, exposing the inner "cat-eyes" structure of the epidermis. Based on Figure 2c, after bleaching, large sheets were absent, and fibrils were visible. Some remnants of large sheets suggest lingering lignin, supported by FTIR. Additionally, segments of fibril sheets implied MFC network agglomeration (Choi et al., 2019).

Figure 2 SEM images with 100x (left) and 500x (right) magnification of each sample; (a) RH, (b) RH_Al (c) RH_Al-Bl (d) RH_Al-Bl-Ch, (e) RH_Al-Bl-Mc

Figures 2d (chemical) and 2e (mechanical) showed that surface patterns, namely RH_Al-Bl-Ch had spiky surfaces, while RH_Al-Bl-Mc was smoother (Galera Manzano et al., 2021). These methods aimed to remove cellulose amorphous phase, thereby enhancing water vapor retention and exposing hydroxyl and carboxyl groups. Sulfuric acid hydrolysis, particularly in the amorphous phase lacking protective steric hindrance in the crystalline region, results in half-ester sulfate groups on cellulose surface, generating a highly negative charge (Rajinipriya et al., 2018). Achieved either chemically or mechanically, cellulose fibrillation divides MFC into smaller fibrils, exposing hydroxyl and carboxyl groups and enhancing water vapor retention in both samples (Väisänen et al., 2018).

3.3.  Chemical Structure Analysis

The chemical structure of the samples was observed through FTIR spectra (Figure 3a). The peak at approximately 1555 cm^-1, attributed to lignin according to Wyman et al. (2004), diminished as expected in RH_Al sample compared to RH, and it became broader and disappeared in samples RH_Al-Bl.

A similar pattern was observed for the slight bump around 800 cm^-1, showing silica in RH samples, which was removed in RH_Al, RH_Al-Bl, and RH_Al-Bl-Ch or RH_Al-Bl-Mc. The valley of 1260 cm^-1 in FTIR signal, represented another impurity, such as hemicellulose, coexisting with the prominent cellulose peak at 1300 cm^-1. This valley weakened and split from RH_Al to further treated samples, signifying the effective removal of silica during the initial alkalization.

Figure 3 a) FTIR spectra absorbance; b) Absorbance ratio of hemicellulose, lignin, silica.

Figure 3b compiles the impurity content in the RH sample, using absorbance ratios to provide a detailed assessment of impurities. The ratios at 800 cm^-1, 1260 cm^-1, and 1555 cm^-1 represent silica, hemicellulose, and lignin, respectively, with the as-received RH sample serving as the reference. Figure 3b shows a significant reduction in all impurities, at least 45%, starting from alkalization treatment. Alkalization substantially reduced silica content, with the entire elimination being performed through bleaching. Hemicellulose and lignin ratios also decreased after alkalization and further decreased due to mechanical or chemical treatment.

Chemical or mechanical treatments were applied to further reduce impurities and extract MFC. The primary objective is to remove the amorphous region and unbundle micro and nanofibrils of cellulose network using different strategies. Chemical treatment uses sulfuric acid, following the same principle as alkalization, to isolate and fibrillate cellulose by hydrolyzing the glucoside bonds in lignin and hemicellulose chains (Jamalpoor and Hosseini 2015).

3.4.  Crystallinity Behavior

3.4.1. Hidrogen bond energy (EH) and bond distance (R)

        In discussing sample crystallization behavior, FTIR absorption peaks in the 1700-850 cm^-1 range aid in assessing cellulose polymorphism. A shift in C-C and C-O vibration peaks showed the conversion of cellulose allomorph from type I to II. According to Carrillo et al. (2004), when cellulose type I is more dominant than type II, the absorption at 1420 and 1155 cm^-1 would change to approximately 1430 and 1162 cm^-1, respectively. Therefore, the absorption at approximately 1430 and 1162 cm^-1 in all sample spectra showed the presence of cellulose I.

        Absorption at 893 cm^-1 for cellulose type I changed to approximately 897 cm^-1. The 895-905 cm^-1 range was in line with higher wavenumbers indicating cellulose type I, consistent with Yue (2011), and attributed to glucose residue rotation around the glycosidic bond (Jamalpoor and Hosseini 2015). Furthermore, all fibers had an absorption peak at 1315 cm^-1, confirming the minor presence of cellulose type II. FTIR analysis reinforces cellulose I dominance in both untreated and treated RH samples.

        EH and R analysis offered another method to distinguish cellulose types. Cellulose type I had 3 hydrogen bonding arrangements, namely intramolecular hydrogen bonds O(6)H···O(2) and O(3)H···O(5), as well as intermolecular hydrogen bond O(6)H···O’(3) with FTIR absorption peaks around 3455-3410 cm^-1, 3375-3340 cm^-1, and 3310-3230 cm^-1, respectively. Meanwhile, cellulose type II had 4 hydrogen bond arrangement, similar to type I, with the addition of an intermolecular hydrogen bond: O(2)H···O’(2) or O(6)H···O’(3) (Poletto et al., 2011). Table 1 provides EH and R values for each sample. Compared to EH value of the as-received RH, all treated samples had lower hydrogen bond energy, consistent with Poletto et al., (2011).

Table 1 Values of EH and R for each sample

        The relative hydrogen bond distances examined in this study range from 2.755 to 2.815 Å. According to Tasker et al., the hydrogen bond distance of Cellulose I was 2.75 Å. Consequently, the proximity of the distance to this value showed the refinement of cellulose in the sample (O’Sullivan 1997). Additionally, as stated by Poletto et al., (2011), higher crystallinity leads to denser cellulose packing and shorter distances, resulting in increased hydrogen energy between cellulose molecules.

        Nelson and O’Connor (1964) proposed the ratio between FTIR absorbance bands at 1372 and 2900 cm^-1 as CrR, which is proportional to the crystallinity degree of cellulose. CrR values for RH, RH_Al, RH_Al-Bl, RH_Al-Bl-Ch, and RH_Al-Bl-Mc were 0.79, 1.73, 2.17, 3.03, and 5.67, respectively. The values gradually increase along with the treatment process, with RH_Al-Bl-Mc having the highest. This suggested that mechanical treatment was more powerful compared to chemical treatment in removing the amorphous region.

3.4.2. Crystallinity Index (CI)

    XRD was used to complement the other characterization and further evaluate the crystallinity behavior of cellulose. Figure 8 shows XRD diffraction of untreated and treated RH. From XRD diffraction data, several peaks were identified, where peak one at =14.5° is  (100) and  (10), peak two at =17.0° is  (010) and  (110), peak three at =22.5° is  or (002) and  (012), as well as peak four at =34.0° is  and  (004) (Hult et al., 2003; Wada and Okano, 2001). XRD peaks were processed by Gaussian deconvolution separation to further determine the crystallinity index, the d-spacing, crystallite size, and Z-values.

In this study, Z-values function developed by Wada and Okano (2001) was used to determine cellulose crystal structure, identifying it as either triclinic  or monoclinic  allomorph. The most intense crystalline peak was discovered at 22.62° on the (002) lattice plane for all the samples. Lionetto et al. and Harahap et al. investigations also identified the same peak within different scanning angles () (Harahap et al., 2023; Lionetto et al., 2012).