Introduction

Lignocellulose derived from industrial and agricultural waste is a renewable material with a potential conversion to several high value compounds. This material can be successfully applied to bio-based production chain types in the circular bioeconomy system, in which reusing and recycling biomass waste is considered to minimize environmental problems related to its accumulation. The bio-waste from agricultural crops and the forestry industry is considerable at the global level, amounting to several gigatons per year. The polymeric constitution of cellulose, hemicellulose, and lignin in their fibers has shown promise as feedstock in several industrial sectors (Oliveira et al. 2016; Pleissner et al. 2016; Menezes et al. 2017; Pires et al. 2018; Meneses et al. 2020). These biopolymers are widely used in traditional industries, such as in the manufacture of paper, chemical, fiber, films, and polymers, but they have also been explored in the production of cellulose at nanoscale (nanocellulose) for applications to new materials with good mechanical properties (Pires et al. 2018; Deepa et al. 2019). Currently, the nanocellulose market has expanded with the construction of industrial plants for commercial production in several countries, such as Canada, the United States, the United Kingdom, Norway, Sweden, Finland, and Japan, thus changing the scope of this market, which was previously restricted to a few countries (Balea et al. 2020).

Among various agricultural crops, citrus fruit, such as the orange, has generated large amounts of residue worldwide. Brazil is currently the largest exporter of orange juice, and orange production is estimated to be 16.8 million tons by 2028 or 2029. This increase will also reflect the amount of bagasse orange waste available as a renewable source of value-added products that have several commercial applications, such as nanocellulose. However, to obtain nanocellulose from any lignocellulosic residue, it is necessary to depolymerize the cell walls by chemical, chemical–mechanical, or even biological treatments (Bhatia et al. 2020). The choice of the raw material and the treatment type depend on the application intended for the cellulose polymer because depolymerization techniques change the size, order, and structure of the cellulose fiber chain, thus changing its morphology (Kakroodi et al. 2014). The depolymerization of cellulose fiber leads to the formation of cellulose nanofibers (CNFs) and cellulose nanocrystals (CNCs). CNFs consist of aggregates of long thread-like bundles of cellulose chain molecules with long, flexible, and entangled, whereas CNCs are rod-like or needle-like particles with high purity, high crystallinity, and high specific surface areas (Khalil et al. 2014; Golmohammadi et al. 2017; Trache et al. 2017; Deepa et al. 2019).

The conventional procedures used in the production of CNFs are mechanical and/or acid hydrolysis or alkali pretreatment, which allow for obtaining flexible nanofibers 5–100 nm in diameter and 500–2000 nm in length (Klemm et al. 2011; Khalil et al. 2012; Brinchi et al. 2013; Deepa et al. 2015; Tan et al. 2015; Phanthong et al. 2018). Among lignocellulosic residues, peel and bagasse orange are promising sources of cellulose nanofibers (Tsukamoto et al. 2013; Hideno et al. 2014; Mariño et al. 2018), which has led to patents (Tasic et al. 2013) and companies that manufacture sustainable fabrics from citrus juice byproducts (Santanocito and Arena 2019). However, the process used to obtain cellulose nanofibers is still conventional (i.e., mechanical, acid, and alkaline) and contrary to the circular bioeconomy.

In this context, ionic liquids (LI) are an alternative to lignocellulose treatment because of their high specificity for the dissolution of amorphous cellulose and their potential for recycling and reusing after the extraction process, thus minimizing negative effects on the environment. In addition, LIs are considered green solvents because they contribute characteristics such as low vapor pressure, high thermal and chemical stability, non-flammability, and the possibility of modulation (Yoo et al. 2017; Halder et al. 2019; Abushammala and Mao 2020). In addition, ionic liquids composed of imidazole salts, have a short cationic chain and, consequently, less steric impedance and greater diffusion capacity, thus increasing the efficiency of the depolymerization process (Zhang et al. 2013; Meenatchi et al. 2017).

CNFs obtained from renewable materials and ecofriendly processes are relevant because they are used as mechanical reinforcement materials in fields such as electronics, food, textiles, medicine, pharmaceuticals, biosensors, and polymers (Dufresne 2017; Rajinipriya et al. 2018; Joshi and Adak 2019; Quesada-González et al. 2019; Lopez-Polo et al. 2020; Wang et al. 2020). In the field of polymer science, biodegradable membranes reinforced with CNFs appear to be a promising alternative in the packaging industry. The use of non-renewable resources is limited because of the harmful effects of synthetic polymers on environmental pollution (Abral et al. 2020; Hai et al. 2020).

Cellulose nanofibers have been used in composites to reinforce water-soluble biopolymers such as starch (Balakrishnan et al. 2017; Yousefhashemi et al. 2019), cellulose acetate (Battirola et al. 2017) and gelatin (Liu et al. 2018; Seo et al. 2020). The materials resulting from this incorporation have been shown to improve the performance of mechanics, porosity, hydrophilicity, and biocompatibility compared with natural matrix polymers. In this context, this paper presents a proposal to synthesize nanocomposites associated with CNFs previously extracted through hydrolysis using ionic liquids and to test their efficiency in mechanical reinforcement.

In this work, the effects of CNF incorporation on starch membrane performance were examined. The study focused on the use of ionic liquids to obtain CNFs from orange bagasse, a waste produced in large quantities, which is explored as a renewable source to obtain value-added products.

Materials and methods

Materials

The raw material used in this work was obtained from orange bagasse (Citrus sinensis) supplied by the Maratá Sucos do Nordeste Ltda factory, which is located in the municipality of Estância/Sergipe. Orange peel was previously dried at 90 °C for 24 h and then ground in a domestic multiprocessor to obtain 32-mesh granulometry. The 1-methylimidazolium ionic liquid was provided by Sigma-Aldrich (USA). All other chemicals used in this experiment were of analytical grade, and the solutions were prepared with deionized water.

Biomass treatment and cellulose nanofiber extraction

Orange residue at 32-mesh granulometry was treated with 1-methylimidazolium ionic liquid at the ratio of 1:10 m/v (residue/IL) and vigorously stirred (1000 rpm) at 90 °C for 2 h. After the treatment, the resulting suspension was subjected to centrifugation at 10000×g for 10 min. Subsequently, the precipitated material underwent a dialysis process in water to remove excess LI (pH between 6 and 7). The suspension was then sonicated for 5 min and stored at 8 °C after the addition of five drops of chloroform (Dong et al. 1998; Pelissari et al. 2014). The residues generated during the biomass treatment and the obtained CNFs (liquid and solid residues) were recovered through the rotary evaporation and reused in other research work.

Preparation of starch membrane reinforced with cellulose nanofibers

Two membranes were prepared using the solvent casting method. The membranes (1 and 2) were prepared by mixing soluble corn starch (1.2 g) with 0.8 g of glycerol and 40 mL of deionized water, which was then subjected to agitation at 1200 rpm for 2 h under a heating bath. The temperatures of the heating bath were 70 °C to prepare membrane 1 and 90 °C to prepare membrane 2. In the last 10 min of both shaking membrane base solutions, 5% (v/v) of cellulose nanofiber solution (11.33 mg/mL) was incorporated.

The nanocomposite solutions were subjected to ultrasound treatment to remove bubbles formed during heating and stirring (Potency: 42 kHz for 10 min). Then 20 mL of the membrane-forming mixture was added to PE plates (polyethylene) and air dried for approximately 48 h at 40 °C. Next, the membranes (nanocomposites) were kept in a desiccator to control their humidity before being characterized (Silva et al. 2015). A membrane without the addition of cellulose nanofibers was used as a control.

Characterization of cellulose nanofibers and nanocomposites

Electron microscopy of cellulose nanofibers and nanocomposites

Cellulose nanofibers and nanocomposites were evaluated by scanning electron microscopy (Shimadzu Super Scan SSX550 model). The membranes without CNFs and nanocomposites were previously lyophilized for 24 h at −55 °C using Liotop’s L101 equipment and later metalized with a thin layer of gold by sputtering at an exposure time of 2 min. The accelerating voltage was 15 kV.

In addition to the SEM, the cellulose nanofibers were evaluated by atomic force microscopy performed on a multimodal AFM (DI, Veeco, Instrumentation Group). Silicon nitride cantilevers (SiNi, silicon nitride) were used in contact mode at a resonance frequency of 1 kHz, a spring constant between 0.15 and 0.30 N, and a peak radius of 100 nm. The samples were prepared by dilution at 0.5 gL−1 in ultrapure water (Milli-Q) and placed on a mica substrate and dried at room temperature. The CNF image analysis was performed using Nanoscope Software 7.3.

Fourier transform infrared spectroscopy (FTIR)

The in-nature orange bagasse residue, as well as the CNFs and the nanocomposite, were submitted to FTIR analysis in ATR mode using Fourier transform infrared spectrometer equipment (FTIR—Nicolet IS10, Thermo Scientific). The spectrum of each sample was analyzed at a range of 4000–500 cm−1; 32 scans per each sample were taken.

Thermogravimetric analysis (TGA)

Thermogravimetric analyses of the samples (8 mg) were performed in a thermal analyzer (TA Q500, TA Company) using a platinum crucible, a nitrogen flow of 20 mL min−1 at a heating rate of 10 ºC min−1 in the temperature range of 25–1000 °C (Chen et al. 2020).

X-ray diffraction (XRD)

The crystallinity of the samples was analyzed using the Shimadzu X-ray diffractometer, model XRD-6000 at CuKα radiation, λ = 1.5406 Å, and an angular pitch of 0.05°. The samples were analyzed in the angular range of 5°–40° (Bragg angle = 2θ) at a temperature of 23 ± 2 °C and an exposure time of 1 s per angle (Ninomiya et al. 2018; Segal et al. 1959). The crystallinity of the samples was determined by the following equation:

$$CrI = \frac{{I_{002} - I_{am} }}{{I_{002} }} \times 100$$

where I002 and Iam are the maximum intensity of the (002) lattice diffraction and the minimum intensity between the (101) and (002) lattice planes, respectively.

Nanocomposite thickness (T) and Tensile test

The thickness of the nanocomposites at 60% RH and 25 °C was measured using a flat-tip Mitutoyo digital micrometer (1 μm resolution) by 3-point random measurements.

The tensile tests were performed in a universal test machine (Discovery HR-3/Hybrid Rheometer, TA Company) with a maximum load of 20 N and a testing speed of 12.5 mm min−1 at 25 °C, according to ASTM D-882. Traction tests were performed on three specimens (15 mm × 2 mm) of each sample (Silva et al. 2015).

Results and discussion

Morphological analysis of cellulose nanofibers

The morphology of the cellulose nanofibers obtained from the 1-methylimidazolium treatment was evaluated by SEM image, as shown in Fig. 1. The lowest magnification image of the cellulose nanofibers is shown in Fig. 1a. SEM image with 30,000 magnification shows a nanofiber tangle formed from the defibrillation process of the sample (Fig. 1b). The CNFs extracted by the treatment with LI is shown in Fig. 1c, which exhibits a network of nanofibers that are arranged in a random manner at diameters not exceeding 100 nm. A similar morphology was observed in the nanofibers obtained by treating with ionic liquids in other residues (Pereira et al. 2020; Moyer et al. 2018; Ninomiya et al. 2018). This result indicated that the use of ionic liquids is an alternative method for obtaining CNF. In addition to scanning electron microscopy, the nanofibers were evaluated by atomic force microscopy (Fig. 2). Thus, it was possible to complement the results of the SEM, confirming that the nanofibers were within the nanometric scale at an average below 100 nm, which was necessary for the synthesis of the proposed nanobiocomposite.

Fig. 1
figure 1

Scanning electron microscopy of cellulose nanofibers extracted from ionic liquid 1- methylimidazolium treatment. The micrographs were captured in increases of 150 × (a), 30,000 × (b), and 300,000 × (c)

Full size image
Fig. 2
figure 2

Three-dimensional graph obtained by the AFM analysis of cellulose nanofibers extracted from orange bagasse using 1-methylimidazolium

Full size image

Thermogravimetric analysis (TGA)

The thermogravimetric analysis performed using natural orange residue, CNFs, and the membrane samples (Fig. 3) revealed four stages of degradation (1º—25–120 °C [approximately]; 2°—120–300 °C; 3°—300–375 °C; and 4°—375–670 °C).

Fig. 3
figure 3

Thermogravimetric analysis of the natural orange bagasse samples, the cellulose nanofibers (CNFs) extracted using 1-methylimidazolium, and membranes with and without the addition of CNFs

Full size image

The first stage was related to the loss of uncontaminated water by dehydration, which was not evident because of the adjustment of the isothermal parameter in the equipment used, which counted the mass loss from the pre-set 40 °C. In the second stage, it was possible to observe the loss of water that was molecularly bound to the structure of the studied samples, as well as the degradation of associated compounds that are on the surface of each material.

The highest percentage of mass loss occurred between the second and third stages, which confirmed the higher concentration of easily degraded polymers and associated compounds (i.e., starch and cellulose). In particular, in the third stage, it was possible to identify the region of the mass loss of cellulose and carbonic chains at the temperature of 375 °C. In addition, in observing the pattern of events, it was identified that the sample of the natural residue exhibited a different pattern of the number of mass loss events. This was evident when it presented a very expressive fourth event between 375 and 670 °C, which was characterized by the degradation of stronger components, such as lignin, which were still associated in the sample.

In general, the technique was aimed to confirm the efficiency of the chemical treatment in removing the other lignocellulosic components (i.e., hemicellulose and lignin), either entirely or partially. This was evident in the thermal behavior of the CNF sample obtained after the extraction process, which showed a single degradation event between 300 and 375 °C, characterized by cellulose degradation, as reported by Meenatchi et al. (2017). At this point, the cellulose nanofibers showed degradation events between 290 and 375 °C (Orrabalis et al. 2019), as shown in Fig. 3. Therefore, no significant difference was observed between the thermal degradation of cellulose results in the literature and the TGA analysis of CNFs. Thus, the thermal degradation of the samples shown in Fig. 3 suggests that CNFs were released during the depolymerization of the lignocellulosic residue.

Fourier transform infrared (FTIR) analysis

The FTIR analysis allowed the identification of the main chemical groups in the evaluated samples (i.e., natural orange residue, CNFs, and membranes) as well as variations in their molecular composition and the presence of CNF in the starch matrix (Fig. 4).

Fig. 4
figure 4

FTIR spectra of cellulose nanofibers (CNFs); starch, Membrane 1 (control); Membrane 1 + CNFs; Membrane 2 (control); Membrane 2 + CNFs

Full size image

The FTIR region between 3000 and 3600 cm−1 corresponded to the hydroxyl group vibration elongation (O–H), which has been associated with the presence of water in analyzed residues (Liu et al. 2018). Peaks in the region of 2900 cm−1 were attributed to the symmetrical and asymmetrical overlap of C–H and the elongation of the vibration of the aliphatic chain in the spectrum, which may be attributed to the cellulose in the sample (Asad et al. 2018; Liu et al. 2018). Thus, it was observed that in the membranes with cellulose nanofibers, the bonds had greater intensity, suggesting the stability of the nanofibers in the membrane matrix.

The bands identified between 1645 and 1632 cm−1 were attributed to the elongated vibration of absorbed water (H–O–H) by carbohydrates with an increasing peak due to less crystallinity (Lee et al. 2015; Rambabu et al. 2016). These bands were also observed in membranes 1 and 2, suggesting a decrease in crystallinity due to water absorption by the starch.

The bands observed near 1428 cm−1 were related to the flexion of C–H present in the carbon chains (Wang et al. 2017), which were related to the concentration of the cellulose chains in the samples. According to Benini et al. (2018), the region close to 1000 cm−1 is associated with the narrowing of CO bonds, which form bands corresponding to the spectra of membranes associated with CNFs. These bands were more evident in membranes 1 and 2 associated with CNFs, suggesting its incorporation during the nanocomposite preparation. In addition, this region was characterized by vibrations related to β-glycosidic bonds. In this case, peaks were observed in the spectra of the membrane samples, especially those with incorporated nanofibers. Their composition included a higher concentration of carbohydrates with this type of chemical bond, which also supported the incorporation of cellulose nanofibers in the formulated membranes.

X-ray diffraction analysis of CNFs and starch membranes

The results of the XRD showed differences between the standard membranes and the membranes with the addition of CNFs, which were evidenced by the intensity of the peaks (Fig. 5). Crystaline peaks near the angles of 14.8°, 16.5°, 22.2°, and 35° corresponded to planes of crystalline cellulose type I. Despite the low intensity of the peak at 35° in relation to the other peaks, it was the most evident in the diffractograms of the membranes associated with cellulose nanofiber (Zain et al. 2014; Meenatchi et al. 2017; Sofla et al. 2016; Mariño et al. 2018). Authors as Kalita et al. (2015), Nascimento et al. (2015) and Asad et al. (2018) also identified peaks within the same regions that were reported in this study, which supports the incorporation of nanomaterial in the polymeric membrane matrix. The peak of 22.2°, which is highlighted in the literature as the reference peak of cellulose nanofibers, was observed in all diffractograms, but at different intensities, thus, indicating the presence of CNFs incorporated in the matrix of the membranes (Miranda et al. 2015; Theivasanthi et al. 2018). This was observed in the difference between the membranes, where those with CNFs in their composition presented higher intensity in all peaks corresponding to the crystallinity of the cellulose.

Fig. 5
figure 5

Diffractogram (XRD) of the cellulose nanofibers (CNFs) extracted using 1-methylimidazolium and membranes with and without the addition of CNFs

Full size image

Moreover, it was identified that the peaks presented a higher base thickness, which characterizes polymorphic material as well as the alternating crystalline and semicrystalline domains of cellulose and starch (Frost et al. 2009; Li et al. 2015).

After identifying the peaks that corresponded to cellulose, the crystallinity index (CI) of the membranes was calculated to highlight increases in the crystallinity of the composite, which were associated with the addition of CNFs in the formulation. The membranes with CNFs showed a CI with values of 16.77% in membrane 1 + CNFs and an IC of 3.13% in membrane 2 + CNFs. In contrast, the membranes of the control group had an IC of 6.89% (membrane 1), whereas a negative IC (-18.51%) was identified in membrane 2 (Table S1).

In general, the low results of the IC of the membranes compared with CNFs (75.23%) were directly related to the low concentration of CNFs in the starch, the major fraction of the matrix, which exhibited a semicrystalline profile (Thiré et al. 2005; Cheng et al. 2017; Pozo et al. 2018). In addition, the negative or null value of CI in membrane 2 (Table S1) was due to the temperature used in the solubilization of the components (90 °C). Temperatures between 90 and 180 °C lead to a change in the molecular organization of the starch, which promote a transition from semicrystalline to amorphous, forming a type of starch known as thermoplastic starch (Blanshard 1987; Corradini et al. 2007; Zhang et al. 2014;Osorio et al. 2019). Therefore, the increase in the amorphous portion of the starch in relation to the crystalline support the result found in membrane 2.

Scanning electron microscopy (SEM) of membranes

According to the results observed in the micrographs of the membranes, it was possible to identify some differences between those that had the starch + glycerol matrix and those that, in addition to these components, had the addition of cellulose nanofibers.

Figure 6 shows the membranes with and without the cellulose nanofiber addition. In Fig. 6a, membrane 1 without CNFs, the surface presents wrinkles and a less homogeneous appearance, while in membrane 2 (Fig. 6c) these irregularities present in the form of depressions. These results could be associated with the preparation methodology and the temperature differences used in their production. In membrane 1, the temperature was 70 °C; in membrane 2, the temperature was 90 °C. Thus, the temperature difference directly influenced the solubilization of the suspension starch (Faria et al. 2017; Souza et al. 2015).

Fig. 6
figure 6

SEM of the membranes produced using 1.2 g of starch, 0.8 g of plasticizer, and 40 mL of H2O through heating in a water bath at 70 °C (Membrane 1 (a); Membrane 1 + CNFs (b)) and 90 °C (Membrane 2 (c); Membrane 2 + CNFs (d))

Full size image

Figure 6b, 6d show membranes with the addition of CNFs, which appeared similar to the others with the addition of a greater concentration of surface irregularities attached to the networks formed by the fibers, which promoted the formation of internal pores that caused irregularities in the surface.

Thickness and tensile testing

As shown in Table 1, the values of the thickness of the membranes synthesized were similar with low standard deviations, which indicated that the tensile tests revealed little relative interference in structural differences, which supported the validation of the data in this study.

Table 1 Thickness, width, and length tests of membranes with and without the addition of cellulose nanofibers (CNFs).
Full size table

Figure 7 a, b show membranes 1 and 2, for which different temperatures were used in preparing the suspension. Therefore, the tensile behavior of the membranes was observed without the addition of cellulose. The axial strengths of membranes 1 and 2 were 0.4 and 0.45 (N/mm2), respectively. The lower GPa represents the modulus of elasticity in membranes 1 and 2, the durations of which were less during the events of elastic and plastic deformation, at 220 s and 170 s, respectively.

Fig. 7
figure 7

Graphs of the tensile test of the membranes. (a) Membrane 1; (b) Membrane 2; (c) Membrane 1 + cellulose nanofibers (CNFs); (d) Membrane 2 + cellulose nanofibers (CNFs)

Full size image

As shown in Figs. 7c and 6d (membranes 1 and 2, respectively, with CNF addition), an increase in axial force (F) was observed at 0.74 and 0.78 N (N/mm2), which corresponded to an increase in strength in relation to the control membrane (without CNFs): 85% in membrane 1, and 73.33% in membrane 2. In addition, an increase in the modulus of elasticity (GPa) was also identified. In membrane 1, the GPa value increased from 2125 to 3500 µm with the addition of CNFs, resulting in an overall increase of 64.7%. In membrane 2, the GPa value increased from 1562.5 to 2312.5 µm with the addition of CNFs, resulting in an overall increase of 48%. The highest values in the tensile test observed in the membranes with CNF addition may be attributed to the improved dispersion in the starch matrix and the compatibility and formation of hydrogen bonds between the starch and CNFs (Agustin et al. 2014; Fan et al. 2016). These physicochemical characteristics can be influenced by the temperature used in the membranes prepared, which explains the high value in membrane 1 with CNFs. According to the literature, the larger the module, the greater the mechanical strength and the resistance to elastic deformation (Berglund et al. 2016; Nechyporchuk et al. 2018). A considerable increase was observed in the limit and time of elastic and plastic deformation at 350 s and 275 s, respectively, in membranes 1 and 2, as well as an increase of cellulose, which corresponded to increases of 59.09% and 61.76% in deformation time, which indicated that more work was needed to promote the rupture of the membranes.

Conclusion

Cellulose nanofibers obtained from orange peel treated with 1-methylimidazolium was used to reinforce starch membrane. The ionic liquid was effective in lignocellulosic matter depolymerization and the consequent release of cellulose nanofibers. The SEM and AFM analysis confirmed CNFs obtention at a diameter not greater than 100 nm, which indicated their potential application in the synthesis of nanocomposites for different purposes. Cellulose nanofibers were incorporated into starch polymer matrices to form a nanocomposite with improved mechanical properties. The membranes with the addition of cellulose nanofibers showed considerable increases in the strength and deformation time parameters. However, membrane 1, prepared at 70 °C, showed the most significant gain in tensile strength (11.7% higher than membrane 2), probably due to the better dispersion of the starch matrix at this temperature in association with the incorporation of cellulose nanofibers. The CNF extraction, as well as the prepared composites, indicated the potential for cleaner alternative processes for obtaining new materials, thus adding value to orange bagasse, a common waste in the agroindustry.