**Recent Trends in the Use of Pectin from Agro-Waste Residues as a Natural-Based Biopolymer for Food Packaging Applications**

#### **Cristina Mellinas, Marina Ramos, Alfonso Jiménez and María Carmen Garrigós \***

Department of Analytical Chemistry, Nutrition & Food Sciences, University of Alicante, ES-03690 Alicante, San Vicente del Raspeig, Spain; cristina.mellinas@ua.es (C.M.); marina.ramos@ua.es (M.R.); alfjimenez@ua.es (A.J.)

**\*** Correspondence: mc.garrigos@ua.es

Received: 31 December 2019; Accepted: 28 January 2020; Published: 3 February 2020

**Abstract:** Regardless of the considerable progress in properties and versatility of synthetic polymers, their low biodegradability and lack of environmentally-friendly character remains a critical issue. Pectin is a natural-based polysaccharide contained in the cell walls of many plants allowing their growth and cell extension. This biopolymer can be extracted from plants and isolated as a bioplastic material with different applications, including food packaging. This review aims to present the latest research results regarding pectin, including the structure, different types, natural sources and potential use in several sectors, particularly in food packaging materials. Many researchers are currently working on a multitude of food and beverage industry applications related to pectin as well as combinations with other biopolymers to improve some key properties, such as antioxidant/antimicrobial performance and flexibility to obtain films. All these advances are covered in this review.

**Keywords:** pectin; food packaging; active compounds; agro-waste residues; circular economy

#### **1. Introduction**

Biopolymers are gaining their market share in the plastics industry by their intrinsic biodegradable character combined with interesting properties for specific applications. Biopolymers can be obtained/ extracted from natural sources, biosynthesized by living organisms or chemically synthesized from biological materials [1]. In addition, their natural-based origin, i.e., from renewable sources represents a great advantage over plastic commodities since their use decrease dependence from petroleum while preserving and even improving important material properties. There has been an increasing interest for the use of biopolymers in packaging, medicine, agriculture, and other sectors. Different types of carbohydrates, such as starch and cellulose, as well as other polysaccharides, such as alginates and pectin, as well as their combinations with animal-protein-based biopolymers, such as silk, wood, gelatin, collagen, chitosan/chitin, gums, plant-based proteins and lipids offer the possibility of rendering interesting applications for these advanced sectors. All these biopolymers offer interesting advantages in their use, such as their renewable origin, biocompatibility, barrier properties to moisture and/or gases, non-toxicity, non-polluting characteristics, mechanical integrity and relative low cost.

The increase in the use of biopolymers has caused that their global market is expected to reach around 10 billion US Dollars by 2021, growing by almost 17% over the forecast period 2017–2021. Western Europe covers the largest market segment, accounting for 41.5% of the global market while other regions are rapidly increasing their market share [2].

In addition, another important possibility offered by the use of biopolymers is their potential to be synthesized from the non-edible parts of plants or animals, avoiding the risk of depleting food from local communities, most of them in under-developed regions. Figure 1 shows the main types of biopolymers that can be obtained from biomass waste as well as some examples of their sources. They can be divided into three large groups: proteins, lipids and polysaccharides. The protein-based biopolymers can be obtained from both animal and vegetable wastes. For example, slaughterhouse wastes are a good source of proteins from animal origin, like gelatin. These wastes comprise the inedible tissues/parts of the animals slaughtered for the production of meat [3]. Among the proteins from plant origin, soy protein isolate is a good option to develop new materials due to its composition and excellent processing ability by gelling, emulsifying ability and water and oil holding capacity [4].

**Figure 1.** Different types of biopolymers obtained from animal and vegetable wastes.

Lipid-based polymers have been used in the last few years in food packaging or 3D printing materials. Their extraction from the natural sources is a necessary step to render isolated fatty acids to be further used in esterification reactions. For example, cutin extracted from tomato by-products [5] and lipids extracted from algae [6] have shown great potential to obtain specific fractions for the production of films with high barrier to water due to the repulsion caused by their high hydrophobic behaviour.

Polysaccharide-based polymers are the last group in Figure 1. These biopolymers are characterized by their biodegradable, biocompostable, sustainable and non-toxic characteristics. Additionally, polysaccharides are more thermally stable than other biopolymers, like lipids and proteins, since they are not irreversibly denatured via heating. However, their main disadvantages are the high sensitivity to moisture and low mechanical resistance [7]. In order to limit these problems, two different approaches have been proposed: the incorporation of different reinforcing additives to the polysaccharides matrices [8,9] and the combination with different polymers to obtain blends [10,11] or multilayer films [12,13]. The improvement in polysaccharides properties have permitted the extension of their use by the food industry in the last few years [14,15].

Pectin is one of the major structural polysaccharides present in many higher plant cells allowing primary cell wall extension and plant growth. It could be extracted and applied as an anionic biopolymer, soluble in water. A large number of recent articles have highlighted the advantages of using pectin over conventional polymers. Therefore, pectin is increasingly important for a multitude of food packaging applications, such as a thickening and gelling agent, colloidal stabilizer, texturizer, and emulsifier [16–18], a coating on fresh and cut fruits or vegetables [19–21] and as micro and nano-encapsulating agent for the controlled release of active principles with different functionalities [22]. Rodsamran et al. [16] reported that bioactive pectin films can retard soybean oil oxidation during 30 days of storage. Furthermore, Sucheta et al. [19] found that a pectin-corn flour-based coating significantly reduced the weight loss and decay per cent of tomatoes, delaying respiration with retention of biochemical quality of tomatoes. Additionally, polymeric blends of hydrocolloids obtained from chia seeds and apple pectin where developed with the aim to obtain antioxidant polymer blend films using the 2,2-diphenyl-1-picrylhydrazyl (DPPH) assay to estimate their antioxidant activity [10].

Pectin has been also extracted from waste biomass by using innovative methods, contributing to waste management in agriculture and food processing industries. Different pectin sources can be used, such as by-products of juice manufacturing as well as orange, mango, banana, lime and pomegranate peels and seeds. Therefore, this review aims to present and discuss the potential of pectin as a bio-based material in food packaging applications by its efficient extraction from waste biomass, while addressing a solution to the important environmental problems caused by the disposal of residues and by-products in the food sector.

#### **2. Pectin**

#### *2.1. Pectin Structure*

Pectin is a complex heteropolysaccharide and a major multifunctional component of the cell wall in many terrestrial plants. It is usually found in association with other compounds like cellulose, lignin or polyphenols present in the cell wall of plants [22]. Pectin is mainly composed of galacturonic acid units (Figure 2). The carboxyl groups of uronic acid residues can be present in different forms in the polymer structure, either free or as a salt form with sodium, calcium or other small counter-ions. In some cases, they can be also present as naturally-esterified groups, particularly with methanol, depending on the pectin source and/or the extraction method. Due to the presence of free carboxyl groups, pectin solutions exhibit acidic pH values. Galacturonic acid comprises approximately 70% of the pectin composition, depending on the plant species, and all the pectic polysaccharides contain galacturonic acid linked at the O-1 and the O-4 positions [23]. Pectin has a linear anionic backbone which regions showing no side chains known as "smooth regions" and regions with non-ionic side chains known as "hairy regions" [24].

Different pectin structural domains may be distinguished (Figure 2), influencing their properties depending on pectin proportions [23].

**Figure 2.** Schematic pectin structure, adapted from [23].

Homogalacturonan (HG): HG is the major domain of pectin in cell walls of plants, representing approximately 65% of the total pectin content. It is formed by galacturonic acid residues, linked by α (1→4) bonds, and their carboxyl groups are partially methyl esterified at position 6. Additionally, this domain may be acetylated at position 2 or 3 depending on the origin of the pectin. These domains are the main constituents of the above-mentioned "smooth regions" [24].

Rhamnogalacturonan I (RG-I) contains about 20%–35% of the total pectin and shows a more complex structure than HG. It contains repeated units of disaccharides consisting of L-rhamnose and galacturonic acid that can be also acetylated in the positions 2 or 3. It could have up to 100 units of (1,2)-α-L-Rha-(1,4)-α-D-GalA. In addition, large amount of L-rhamnose structures are substituted at O-4 by different neutral sugars such as L-galactose and L-arabinose [23].

Rhamnogalacturonan II (RG-II) contains about 10% of the total pectin and it is the structurally more complex component. Despite it is a relatively minor component in the pectin chain, RG-II plays a central role in the structure of plant cell walls. Small structure modifications of RG-II lead to reductions in the dimers formation and they can cause severe growth defects. So, dimerization of RG-II in the cell wall may be crucial for the normal growth and development of plants. This domain is composed of an HG backbone of (at least eight) 1,4-linked α-d-GalA residues decorated with side branches consisting of different types of sugars (rhamnose, fucose, xylose, galactose, apiose or aceric acid) in over 20 different types of linkages [23,25,26].

It is generally believed that the pectic polysaccharides are covalently bonded with high crosslinking densities since harsh chemical treatments or digestion by pectin-degrading enzymes are required to isolate HG, RG-I, and RG-II from each other. In addition, it has been reported that other components, such as xylogacturonan (XGA) and apiogalacturonan (AP), could replace the galacturonic acid units in some parts of the pectin chain [23]. The complexity of the pectin structures increases since it can be changed during the plant storage, extraction and processing, resulting in modifications of pectin functionalities and hindering its structural elucidation. It has been reported that the variations in chain lengths in each of the different domains are not the same, because HG and RGII have a highly homogeneous structure while RGI exhibits a wide heterogeneity in its composition [26].

#### *2.2. Type of Pectins*

The degree of esterification (DE) is an important parameter for the definition of the pectin applications and it is defined as the percentage of carboxyl groups esterified present in the structure of pectin. DE is often used to classify the different types of pectins (Figure 3). Depending on the DE different emulsifying, texturizing and gelling properties are observed. In general, when the DE increases, the water solubility decreases due to the hydrophobic nature of esters with long hydrocarbon chains. In contrast, when the DE increases the gelation rate also improves resulting in rapid gelation pectins [27]. Furthermore, the amount and composition of neutral sugars and the overall molecular weight of pectins have a great influence in their rheological properties [28].

**Figure 3.** Structure of low and high methoxyl pectins.

High methoxyl pectin (HMP) shows DE higher than 50% and it is mainly used in the food industry by its thickening and gelling properties. It has been reported that HMP requires high amount of sugars for gelation and it is very sensitive to acidity [29]. HMP forms gels at low pH values and high concentrations of soluble solids due to the presence of hydrogen bonding and hydrophobic interactions between the pectin chains. Neutral sugars like sucrose play different roles in gelation through regulation of the hydrophobic interaction or directly binding to the polymer chains of HMP. Gels are formed when HG portions are cross-linked to form three dimensional crystalline networks in which water and other solutes are trapped [30]. The mechanism of formation of HMP gels is complex and it has been the subject of many investigations in the last decades. The glass transition theory has been proposed to explain the formation of HMP gels. Due to the high viscosity of molecules an arrest of the system kinetics occurs, giving as a result the formation of the gel due to the increased concentration of co-solutes and decrease in the water content. The transition from sol to gel behaviour is due to the combined effect of HMP and sucrose at pH 3, and occurs when excluded volume effects and attractive interactions are capable to give rise to an incipient three-dimensional network [31]. In addition, the effect of monovalent cations has been evaluated under different alkaline conditions (NaOH and KOH) at different pectin concentrations. It was suggested that HMP gel is formed through different mechanisms, such as de-esterification, self-aggregation, and entanglement under alkaline conditions. Na<sup>+</sup> or K<sup>+</sup> bind to dissociated carboxyl groups in HMP due to electronic attraction and this behaviour enables HMP molecules to move closer to each other, thereby improving gel network formation [32].

The emulsifying properties of HMP have been investigated by Jiang et al. in binary water-ethanol systems. They suggested that ethanol reduces the electrostatic repulsion and promotes pectin aggregation [33]. When HMP was mixed with the water-ethanol mixtures, the helix structure was broken, and electrostatic repulsion decreased. The compact and hydrophobic conformation enables pectin to adsorb better on the oil-water interface. The obtained emulsion showed good stability when using 21% of ethanol in the mixture.

Low methoxyl pectin (LMP) shows DE lower than 50% and it is generally formed by the de-esterification of HMP. Different agents can be used for the preparation of LMP from HMP, such as alkalis like sodium hydroxide or ammonia, enzymes (pectin methyl esterase) and concentrated acids [34].

LMP is widely used by the food industry to form low sugar-content jams as it does not require large amounts of sugar for gelation. It shows less sensitivity towards acidity and requires Ca2<sup>+</sup> ions to form gels. The gelation mechanism in LMP is mediated by the formation of calcium bonds between two carboxyl groups from two chains in close contact [28]. Recently, Han et al. studied the effect of the calcium concentration, pH, soluble solids, and pectin concentration on the gel strength of LMP gels and they proposed different mechanisms of formation of pectin gels based on their rheological properties [35]. They observed that pH values close to the isoelectric point (pH = 3.50) and high calcium concentrations enhanced the storage modulus and gel strength by formation of calcium bridges at dissociated carboxyl groups. In addition, the sucrose content improved the gel strength because the neutral sugars provide hydroxyl groups to stabilize the gel and contribute to the formation of hydrogen bonds to immobilize free water. On the other hand, the formation of LMP gels under alkaline conditions was tested by Yang et al. [36]. They suggested that LMP can form relatively stable gels in the pH range of 3.5–9.5 using NaOH as pH regulator. In addition, they evaluated the effect of calcium concentration in the thermal and structural properties of the LMP gels and they concluded that the presence of calcium ions not only reduced the thermal stability but also the crystalline degree of LMP.

#### **3. Sources of Pectin**

Due to the high potential of pectin-based polymers, the extraction of pectin from biomass waste has been widely studied. Table 1 summarizes several published works based on the extraction of pectin from agro-waste sources.


**Table 1.** Different raw materials and extraction methods to obtain natural pectin.

HAE: Hydrothermal-assisted extraction; UAE: Ultrasound-assisted extraction; HC: Hydrodynamic cavitation; MAE: Microwave-assisted extraction; SWE: Subcritical water extraction; EAE: Enzyme-assisted extraction.

The peels of citrus fruits have been reported as the main source to obtain pectin at the industrial scale due to their good properties and high extraction yield. Hydrothermal extraction is the most usual method to obtain pectin from orange peels and it involves high temperatures (75–95 ◦C) and extraction times (60–300 min). Additionally, in all cases, the hydrothermal extraction of pectin takes place under acidic conditions using water as solvent. Pectin is very soluble in water and the acid medium decrease the presence of other compounds like polyphenols increasing extraction yields and helping to maintain the quality of the extracted pectin [38–41]. Other methods have been tested to reduce extraction times in citrus by-products. For example, microwave-assisted extraction (MAE) has been used in lime [64] and pomelo peels [65] reducing the extraction times to five and two minutes, respectively. However, high microwave powers (700–1100 W) were required to achieve these results. The hydrodynamic cavitation method was also used to obtain pectin derived from orange peel waste. Although a large decrease in the amount of solvent (2.86 mL/g of dry waste) was observed, long extraction times were also needed (270 min) [55].

The use of other sources to obtain pectin-based polymers in good grade and quality has been proposed in the last few years, such as eggplant peel [37], chamomile waste [45], cocoa pod husk [59,66], banana peel [49], mango peel [50,61,67] or tomato husk [28]. Tropical fruits have been also studied in the last years to obtain HMP. For example, passion fruit rind [54,62], durian rind [46] or jackfruit peels [51,52,63] have been proposed as interesting sources of pectins. Hydrothermal extraction is also the most used method in these types of wastes. Ultrasound-assisted extraction (UAE) has been also tested in passion fruit rind using 450 W and a water to dry sample ratio of 20 mL/g for 10 minutes. Results showed that the obtained pectin was mainly formed by homogalacturonans. Furthermore, their high degree of methylation indicated that the passion fruit pectin could be applied in gel forming products [62].

The use of innovative and sustainable extraction techniques is heading towards the study of hybrid techniques with the objective of combining their advantages, such as in the case of MAE and UAE. Pectin has been obtained from sisal waste by the combination of enzymatic and ultrasonic processes as an efficient strategy for the production of high-quality pectins since the enzymatic treatment disrupt the links between cellulose and xyloglucans in the cell wall of sisal and then the ultrasonic treatment produces mechanical destruction of the sisal structure to improve the release of pectin [57].

Finally, the introduction of new extraction techniques can be a great initial investment for companies since they offer the possibilities to get specific extractions of high added value purified compounds, although the costs of microwave or ultrasonic based equipment are higher than those of conventional extraction equipment, but in the long term, these devices are more profitable since the energy consumption, extraction time and the amount of expensive reagents used during pectin extraction are reduced [68,69].

#### **4. Pectin-Based Materials for Food Packaging Applications**

Pectin is a versatile compound that can be used to develop different materials in many food applications such as thickening and gelling agent, colloidal stabilizer, texturizer and emulsifier. These important applications are not limited to food processing, but also to packaging, coatings on fresh and cut fruits or vegetables and as microencapsulating agents (Table 2). Pectin is soluble in pure water and insoluble in organic solvents. Moreover, when dry pectin is mixed with water it tends to hydrate very rapidly, forming clumps. This behaviour is due to the formation of dry spheres of pectin contained in a highly hydrated outer coating. In order to eliminate these clumps, a vigorous and long agitation time is required [70]. In general terms, diluted pectin solutions present a Newtonian behaviour, but at high concentrations they show non-Newtonian behaviour, corresponding to pseudo-plastic characteristics. It was observed that the decrease in solubility and increase in viscosity contribute to increase the gelation capacity, i.e., the pectin concentration has a positive effect in gelation capacity and viscosity but a negative effect in solubility. Although it was stated that pectin properties are mainly

dependent on structure, particularly DE [22,71], film forming, gelling and emulsifying properties should be also considered.


**Table 2.** Different types of pectin-based materials used in food packaging applications.

EO: essential oil; AM: antimicrobial; AO: antioxidant.

#### *4.1. Pectin-Based Films*

Casting is the most used technique to obtain pectin-based films [10,73–75,79]. Pectin solutions (around 2–3 wt%) are mixed with the appropriate amount of plasticizer, commonly glycerol [95]. Then, the film forming solution is dried under controlled conditions of temperature and humidity forming a thin film. The incorporation of active agents, such as antimicrobial and/or antioxidant compounds, is performed after the incorporation of the plasticizer to obtain good compatibility between all components during the film processing [96–98]. Recently, Gouveira et al. [99] have reported the successful production of pectin-based films by using thermo-compression moulding of raw pectin with a natural deep eutectic solvent. The visual aspect of the obtained films was acceptable, since they were yellowish, visually homogenous, semi-transparent and without apparent pores, also showing high tensile strength and water resistance.

Pectin offers good compatibility with other biopolymers, such as proteins [73], lipids [100], other natural polysaccharides [101] or even synthetic biopolymers [82]. All these combinations represent alternatives when considering the final application of the obtained films. Both types of pectin (HMP and LMP) can form thin films under specific conditions. For example, LMP has been used as an appropriate matrix in new antioxidant systems with ascorbic acid as active additive by using casting as the processing method [72]. LMP was heated to 90 ◦C and then, glycerol and ascorbic acid were incorporated to the solution. Finally calcium chloride was added as the crosslinking agent to permit the formation of consistent and homogeneous pectin-based active films. In contrast, the addition of calcium ions is not necessary to develop films based on HMP, but low pH values and high sugar concentrations are needed to produce thin films. Nisar et al. [74] produced HMP films with antimicrobial properties incorporating clove essential oil by the casting method. Film forming solutions (3% *w*/*v*) were prepared by rehydrating pectin in sterile deionized water for 12 h at 20 ◦C. Glycerol was used as plasticizer at 30 wt % with magnetic stirring at 70 ◦C while pH was adjusted to 4.5. The clove essential oil with an emulsifier to improve the oil dispersion in the film aqueous solution were incorporated into the film forming solutions at different concentrations. A great integration of the clove essential oil into the polymer matrix was observed with a positive significant influence on the physico-chemical and functional properties, in particular barrier, mechanical, antioxidant and antimicrobial.

Marjoram, mint and rosemary essential oils are some of the active additives incorporated into the pectin matrix to get functionalities to these biopolymer materials. Almasi et al. [75] evaluated the effect on physico-chemical properties of marjoram essential oil in pectin films for food packaging applications. In order to prevent the degradation of the highly volatile essential oil and to control its release into food, it was incorporated using nanoemulsions and Pickering emulsions. Both types of emulsions combined the use of the essential oil and a low molecular weight surfactant with whey protein isolate or inulin as nanocarriers. Results obtained through X-ray diffraction (XRD), Fourier transformed infrared spectroscopy (FTIR) and field emission scanning electronic microscopy (FESEM) confirmed the high compatibility between pectin and both emulsions. The encapsulation of the essential oil through Pickering emulsions provided significantly slower releasing rates through the films when compared to nanoemulsions. For these reasons, authors concluded that the active pectin films containing Pickering emulsions showed the best potential to be used in active food packaging due to the slow release of the essential oil increasing food shelf-life. On the other hand, the synergic effect between mint and rosemary essential oils and nisin was investigated by Akhter et al. [78] using chitosan, starch and pectin blends. These authors concluded that the inclusion of rosemary essential oil and nisin improved the water barrier properties, tensile strength and thermal stability of the active biocomposites. Additionally, the combination of these compounds in a pectin matrix showed high antimicrobial action against some pathogenic strains (*Bacillus subtilis*, *Escherichia coli* and *Listeria monocytogenes*).

Furthermore, extracts derived from plants have been proposed to improve the functional properties of pectin. For example, tea extracts were incorporated into a HMP/Glucomannan blend. The influence of the addition of tea extracts at different concentrations (from 1% to 5% wt % on a dry basis) on the structural and physical properties of the blend, as well as on the antioxidant and antimicrobial activities were evaluated [76]. The authors found that concentrations of tea extract lower than 2 wt % improved all these properties but the effect was negative at high concentrations since some aggregation in the biopolymer macromolecules was observed. Red cabbage extract has been proposed for the development of a smart film based on HMP for meat and fish products. [80]. Red cabbage extract is rich in anthocyanins, showing the ability to change the colour of the biopolymer matrix at different pH values. It is known that the degradation of animal proteins produces an increase in pH due to the liberation of nitrogen compounds that can be monitored using a colorimetric sensor based on pectin and the red cabbage extract, offering innovative applications of pectin to the food industry [80]. These results showed the significant colour change in edible films when they are exposed to the headspace of meat and fish products at 21 ◦C and 4 ◦C, respectively.

The physical and functional properties of pectin-based films can be also modified by the combination of commercial pectin with corn flour and beetroot powder to minimize post-harvest decay, reducing ripening and improving sensorial properties of tomatoes [19]. In this study, results showed that pectin-based films protect from losses of polyphenols improving the antioxidant activity of these materials. In addition, other properties are modified due the presence of the edible coating. Pectin can modify the atmosphere around the fruit and/or vegetables, altering oxygen levels inside the fruit, retarding production of ethylene and, thus, limiting their physiological decay. In this work,

pectin-based films showed low hydrophobicity to get optimum gas and water vapour permeability; reducing the ripening induced quality degradation in terms of texture and loss of bioactive compounds during storage.

Finally, the incorporation of nanoparticles to improve the physical and functional properties of pectin-based films has been evaluated. Biocomposites formed by a biopolymer matrix with metal or metal oxide nanoparticles are gaining importance in active food packaging since they could play a double role. On one hand, nanoparticles can act as nanofillers to enhance the mechanical and barrier properties of the biopolymer matrix and, on the other hand, they can interact directly with food due to their potential antimicrobial/antioxidant activity [102]. The effect of silver nanoparticles (AgNPs) has been tested in pectin, pullulan (a polysaccharide produced by fermentation by the *Aureobasidium pullulans* fungus), and pectin/pullulan blends [77]. Silver nanoparticles improved the mechanical properties of pullulan/AgNPs and pullulan/AgNPs/pectin composites while also showing high antimicrobial activity against foodborne pathogens, especially *Salmonella Typhimurium*, *Escherichia coli* and *Listeria monocytogenes*. AgNPs have been also proposed to develop nanocomposites based on pectin to be used as coatings for other polymer matrices with the aim to improve their barrier and mechanical properties as well as providing antimicrobial/antioxidant properties. Nanocomposites based on pectin with AgNPs and laponite have been evaluated to get a significant reduction in the oxygen transmission rate and water vapour transmission rate respect to neat polypropylene films taken as control [82]. The application of these new films showed antimicrobial activity against Gram-negative and Gram-positive bacteria, *Escherichia coli* and *Staphylococcus aureus*, respectively.

Other types of nanoparticles have been tested in pectin as the polymer matrix. Titanium oxide nanoparticles (TiO2NPs) were incorporated at low concentrations (0–2 wt %) into biodegradable starch–pectin (3:1) films to improve their mechanical and barrier properties as well as their potential as antioxidant systems for food packaging applications [103]. In addition, visible and UV radiation was completely absorbed or scattered in these films by the addition of TiO2NPs to get starch-pectin films with potential as UV screening packaging materials. On the other hand, the addition of halloysite nanotubes (HNT) offers great advantages to develop advanced food packaging materials. The effect of HNTs with salicylic acid [104], rosemary [97] and peppermint [105] essential oils in pectin films has been reported. HNTs showed high compatibility with pectin films improving their mechanical, thermal and moisture barrier properties [97,104]. The antimicrobial performance of these films was also improved due to the increase in the release rate of active compounds [97]. In fact, the antimicrobial activity of pectin-based films against Gram-negative *Escherichia coli* ATCC 25922, *Salmonella Typhimurium* ATCC 14028, *Pseudomonas aeruginosa* ATCC 10145, and Gram-positive *Staphylococcus aureus* ATCC 29213 was studied by the disk diffusion method, suggesting the effective antimicrobial properties of these functionalized films [104].

#### *4.2. Emulsions and Gels*

Pectins are widely used in the food industry as emulsifier and gelling agents. The ability of pectins to form gels under specific conditions has been used to obtain aerogels [86,106], hydrogels [87,107] or oleogels [88,108]. In particular, hydrogels are the most popular gel compositions used in food packaging, since they are able to absorb large amounts of water or other biological fluids inside their structure. For example, Torpol et al. studied the encapsulation of two different antimicrobial compounds: garlic and holy basil essential oils in chitosan-pectin hydrogel beads [87]. The entrapment of essential oils in the matrix structure was successful and it showed antimicrobial capacity against *Bacillus cereus*, *Clostridium perfringens*, *Escherichia coli*, *Pseudomonas fluorescens*, *Listeria monocytogenes* and *Staphylococcus aureus*, but not against *Lactobacillus plantarum* and *Salmonella Typhimurium.* Hydrogel coatings have been also proposed to reduce the deterioration of fresh fruit, meat or fish, as they can provide a semi-permeable protection to gases and water vapour and some other environmental factors that could damage food. By promoting food perspiration, these films also help to reduce enzymatic browning and water loss. Furthermore, this protection may also be enhanced by the addition of other ingredients, such as minerals, antioxidants, nutrients, vitamins or probiotics [109]. On the other hand, when the extraction of solvents at their supercritical state is produced in hydrogels or alcogels, the resultant material is called aerogels. Due to their unique properties, such as high porosity, high specific surface area, low relative density and thermal conductivity, these pectin-based biopolymers represent an innovative approach as advanced materials for food packaging since they can be used as internal layers, oxygen scavengers or drug delivery systems [110,111]. Recently, pectin-based aerogels have been developed for the storage of temperature-sensitive food. In this regard, TiO2 nanoparticles were incorporated into the pectin matrix to improve the mechanical, thermal and antimicrobial properties of pectin when compared to control films [86].

As it has been mentioned above, pectin can be used as nanoemulsions, which are kinetically stable, but thermodynamically unstable, systems whose production requires emulsifiers to stabilize the dispersed phase [93]. Different types of essential oils have been encapsulated using nanoemulsions to delay and control the release processes. Several authors have studied the use of nanoemulsions in pectin matrices with different essential oils extracted from curcumin [91], lemongrass [92,93] and oregano [92]. Although essential oils are especially interesting in food packaging applications due to their antioxidant and/or antimicrobial properties, Mendes et al. [93] incorporated pectin nanoemulsions with the lemongrass essential oil into cassava starch film to improve the biodegradation rate of these formulations. Results obtained for the film with nanoemulsions showed a suitable degradation in vegetal compost, ensuring their complete biodegradation in a short time increasing their potential application in the food industry.

#### **5. Conclusions**

Researchers and scientists have achieved great success in the development of new systems based on pectins, as a natural bio-based biopolymer that can be obtained from agro-waste products, contributing to the implementation of the circular economy concept by improving waste management. This review article has considered the latest results obtained by researchers on the extraction, functionalization and potential applications in the food industry (including packaging), such as the production of films, emulsions and gels. However, due to the variability in the pectin structure the final application of pectin matrices is very diverse but very promising in many fields related to food packaging, particularly when active formulations are searched. Further studies on pectin matrices and optimization of polymer processes will be needed to better control the resulting pectin-based products.

**Author Contributions:** Conceptualization: C.M., M.R., A.J. and M.C.G.; methodology: C.M., M.R., A.J. and M.C.G.; formal analysis, discussion and supervision: C.M., M.R., A.J. and M.C.G. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Acknowledgments:** Authors would like to thank the Spanish Ministry of Science, Innovation and Universities for their support through the project referenced MAT2017-84909-C2-1-R.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **Size Distribution and Characteristics of Chitin Microgels Prepared via Emulsified Reverse-Micelles**

#### **Siriporn Taokaew \*, Mitsumasa Ofuchi and Takaomi Kobayashi**

Department of Materials Science and Technology, School of Engineering, Nagaoka University of Technology, 1603-1, Kamitomioka, Nagaoka, Niigata 940-2188, Japan; technomare3156@gmail.com (M.O.); takaomi@vos.nagaokaut.ac.jp (T.K.)

**\*** Correspondence: t.siriporn@mst.nagaokaut.ac.jp; Tel.: +81-258-47-9383

Received: 11 March 2019; Accepted: 8 April 2019; Published: 10 April 2019

**Abstract:** Chitin was extracted from local snow crab shell waste and used as a raw material in the fabrication of porous spherical microgels. The chitin microgels were obtained using a batch process of emulsification and, afterward, gelation. The effects of chitin concentrations, oil and water phase ratios (O:W), surfactants, and gelation on the size distribution and morphology of the microgels were investigated. The extracted chitin possessed α-chitin with a degree of acetylation of ~60% and crystallinity of 70%, as confirmed by Fourier Transform Infrared Spectroscopy (FTIR) and X-Ray Powder Diffraction (XRD). In the reverse-micellar emulsification, different chitin concentrations in NaOH solution were used as aqueous phases, and n-hexane media containing Span 80-based surfactants were used as dispersion phases. Various HCl solutions were used as gelling agents. Microgels with sizes ranging from ~5–200 μm were obtained relying on these studied parameters. Under the condition of 3% *w*/*w* chitin solution using O:W of 15:1 at 5% *w*/*w* of Span 80 (hydrophilic-lipophilic balance; HLB of 4.3), the gelation in the emulsified reverse micelles was better controlled and capable of forming spherical microgel particles with a size of 7.1 ± 0.3 μm, when 800 μL of 1 M HCl was added. The prepared chitin microgel exhibited macro-pore morphology and swelling behavior sensitive to the acidic pH.

**Keywords:** crab shell; chitin; spherical microgels; reverse micelle; gelation

#### **1. Introduction**

Chitin (*ß*-(1,4)-*N*-acetyl-D-glucosamine) is a natural polysaccharide found in shells of marine animals such as shrimp, lobster, and crabs. As a marine fishery product, about 30 thousand tons of crabs were annually caught in Japan [1]. This means that large amounts of crab shells were disposed in landfills without being recycled. It is known that crab shells are a good biomass source of chitin [2]. Due to biocompatibility and non-allergenicity, chitin has been widely used in pharmaceutics and bio-medical drugs. In addition, chitin extracted from crab shells has other characteristics such as antibacterial properties and protein affinity that are useful for wound dressing and controlled-drug release applications [3,4]. As a drug carrier, small spheres of gel forming chitin have been recognized as having high drug loading capacity, efficient drug control at the target site, sustained drug release, and high stability compared to micelles and lipid-based carriers [5,6]. Moreover, microgels can be applied as adsorbents, chemical/biological sensors, enzyme immobilization, and gene delivery vehicles [5,7].

The applications of chitin and its derivatives as microgels/microparticles were reported as biological filling [8] and drug delivery agents [6]. Chitin microparticles could regulate the depletion of cholesterol by cellular macrophage activation [9]. A fragmented physical hydrogel suspension of chitin derivatives was indicated to support reepithelization of spinal tissue and vasculature with minimal fibrous glial scaring [8]. Moreover, the fragmented chitin microgels loaded with an anti-metabolite drug for delivery in psoriasis treatment exhibited higher skin permeable efficacy than those of the control drug solution and the conventional drug gel. The drug-loaded fragmented chitin microgels also exhibited greater swelling and drug release at acidic pH than in neutral and alkaline conditions [6]. However, the time window of the use of microgels with an unspecific-shape was difficult to determine with high precision because the shape and the stimuli responsiveness influence the biodistribution, the circulation dynamics, the drug release, and the intracellular uptake of the microgels [8,10]. Hence, the fabrication of microgels that have precise geometries and stimuli-responsiveness has been significant in particle transportation and therapeutic agent delivery.

It has been reported that several microgel preparation methods including solid-phase organic synthesis [11], microfluidics [12], and emulsification [13,14] were feasible. However, solid-phase organic synthesis and microfluidics have numerous problems associated with the use of cross-linked insoluble polymers, the fluctuation of reaction rates, and the longer time-consumption in such processes [15,16]. In contrast, a reverse-micellar emulsification technique simplifies the process, making it an effective tool to synthesize small particles with controllable size and shape. An emulsion-based method is also energy-efficient, non-destructive, and attractive for large-scale production [13,14]. As compared to the other approaches, the reverse-micellar emulsification can enhance uniformity and dispersity of the polymeric particles, can be operated at low temperature, and provide a stable dispersion for a water in oil emulsion system [17]. Therefore, such a method is used to prompt self-assembly of surfactant in organic media, whereby the oil region having a nonpolar nature faces the outside surface of the micelle, and the polar region forms the core for polymeric microgels [18]. In such a structure, the tiny aqueous droplets with varied sizes are encapsulated, and the different-sized microgels are produced within the reverse micelle after gelation [13]. Accordingly, chitin microgels with the same size prepared by the reverse-micellar emulsification method can be described.

The aim of the present study was to prepare chitin microgel by using a reverse micelle system at various compositions of water in oil (W/O) emulsions. The chitin used in this study was extracted from shell waste of red snow crabs, which was collected from the local area in Niigata prefecture, Japan. The extracted chitin was then characterized and compared to commercial chitin and chitosan. The synthesis of the microgel was performed using a batch process of W/O emulsion. The effects of chitin concentrations (water phase) of 1–3% *w*/*w* and oil:water phase ratios of 3–15:1 were studied. It was known that hydrophilic-lipophilic balance (HLB) values of surfactant ranging from 3.5 to 6 were more suitable for a W/O emulsion system [19]. For Span 80, a nonionic-based surfactant, the HLB value of 4.3 was adjusted to 5 and 6 in this study. Concentrations of Span 80 (3–7% *w*/*w*) containing n-hexane (oil phase), and gelation using HCl were also investigated in terms of their size distribution and morphology.

#### **2. Materials and Methods**

#### *2.1. Materials*

Dried, cleaned snow crab shells, *Chionoecetes opilio*, were obtained from Teradomari port, Teradomari, Niigata prefecture, Japan. Chemicals were purchased from Wako Pure Chemical Industries, Ltd, Osaka, Japan. Distilled and ion-exchanged water was used in all the experiments.

#### *2.2. Extraction of Chitin*

The coarse flakes of crab shells (30 g) were hydrolyzed using 900 mL of 1.0 M HCl under stirring at room temperature (20 ± 5 ◦C) for 24 h. The reaction was stopped by adding water and filtered through a mesh sieve to remove small contaminants. Protein residuals were removed by heating the hydrolyzed chitin at 90 ◦C in 900 mL of 1.0 M NaOH under stirring for 5 h. Pigments in chitin were removed by stirring in 900 mL of ethanol for 5 h at 60 ◦C. The extracted chitin was dried in vacuum oven at 60 ◦C for 24 h, and ground in a blender.

#### *2.3. Preparation of Microgels*

The chitin powder was dissolved in 20% *w*/*v* NaOH at −20 ◦C under periodic stirring to obtain 1, 2, and 3 % *w*/*w* of chitin aqueous solution. Before emulsion formation, the oil layer solution consisted of n-hexane and surfactants were used as a dispersion phase and prepared in a 50 mL amber vial. According to the critical micelle concentration (CMC) of Span 80 in n-hexane, see Figure 1, below the CMC in the presence of surfactant monomer, there was no peak throughout the spectrum, see Figure 1a, but the peaks at 270 nm appeared at above the CMC. The CMC was approximately 0.25% *w*/*w*, determined by the change of the trend in Figure 1b, which indicates the initial formation of micelles. Above the CMC value (about 10 times), a number of surfactant molecules were able to gather and form stable micelles in the bulk liquid [20]. Therefore, 3, 5, and 7 % *w*/*w* of Span 80 were adopted for this study. A Span 80 (Sorbitan monooleate)-based surfactant was mixed with sodium cholate (HLB 18) to obtain HLB values of 4.3, 5, and 6. The chitin solution was dropped into the dispersion phase with vigorous stirring at 1500 rpm at room temperature (20 ± 5 ◦C) for 45 min. The W/O emulsion was then heated to 65 ◦C. Aqueous HCl solution in the range of 0.01–0.1 M concentration was used as counter-ions. In the gelation process of chitin microgel, 400–1200 μL of aqueous HCl solution was periodically dropped into the emulsion under stirring at 150 rpm. The parameters tested in the preparation of microgels are shown in Table 1. The chitin microgels were coagulated in the liquid medium and precipitated. The microgel was purified to remove the surfactant and residual n-hexane using dialysis (Molecular weight cut-off of 12 kDa, 0.5 nm, AS ONE corporation, Osaka Japan) in 1L of distilled deionized water for 72 h.

**Figure 1.** UV-Visible absorbance versus concentration profile of Span 80 in n-hexane at 20 ± 5 ◦C (**a**), and absorbance at 270 nm of Span 80 at various concentrations (**b**). Span 80 was dissolved in n-hexane at a concentration of 0.01–7% *w*/*w*. After thorough mixing, the solution was transferred to a 1.0 cm quartz cell and the spectrum was recorded at wavelengths of 200–400 nm using UV-visible near-infrared spectrophotometer (Jasco V570, Jasco Corporation, Tokyo, Japan). Blank n-hexane was used as a reference. The vertical dashed line in (**b**) marks the critical micelle concentration.

#### *2.4. Characterization*

#### 2.4.1. X-Ray Fluorescence Spectroscopy (XRF)

μ An elemental study of the extracted chitin was performed using an X-ray fluorescence spectrometer (Rigaku ZSX Primus II, Tokyo, Japan) using ZSX software. This spectrometer contains a 50 keV and 50 mA X-ray tube, providing the detection of diverse elements of the Periodic Table. Prior to characterization, the sample pellets were prepared by using pressed powder method under a pressure of 500 kgf/cm2.

#### 2.4.2. X-Ray Powder Diffraction (XRD)

X-ray diffractograms were obtained using an X-ray diffractometer (Rigaku Smart Lab 3 kW, Tokyo, Japan) under operation conditions of 40 kV and 30 mA with Cu Kα radiation. The relative intensity was recorded in steps of 0.1◦ and at a speed of 3.0 ◦/min. The crystallinity index (*CrI*) was determined by integrated X-ray powder diffraction software (Rigaku PDXL2, Rigaku Corporation, Tokyo, Japan). The quantitative analysis was performed based on the Rietveld refinement and an ab-initio crystal structure determination using crystal structure information of α-chitin provided by the software. The degree of acetylation (*DA*) of chitin [21] was calculated by:

$$DA\left(\%\right) = 100 - \frac{(103.97 - \text{CrI})}{0.7529} \tag{1}$$


**Table 1.** Parameters in preparation of chitin microgels.

#### 2.4.3. Fourier Transform Infrared Spectroscopy (FTIR)

FTIR spectra were obtained using a FTIR spectrometer (Jasco 4100, Jasco Corporation, Tokyo, Japan). The sample pellets were prepared using the KBr method. The absorption bands were scanned between 4000–400 cm<sup>−</sup>1. The degree of acetylation (DA) was calculated by:

$$DA\ \left(\%\right) = \frac{1}{1.33} \left(\frac{A\_{1655}}{A\_{3450}}\right) \times 100\tag{2}$$

where *A*<sup>1625</sup> and *A*<sup>3450</sup> are values of absorbance measured at 1625 and 3450 cm<sup>−</sup>1, respectively [22].

#### 2.4.4. Differential Scanning Calorimetry (DSC)

Thermograms were carried out using differential scanning calorimetry (DSC) (Rigaku, Thermo Plus EVO DSC823, Tokyo, Japan) under an air atmosphere. Dried samples (3–5 mg) were placed in hermetically sealed Al pans and immediately loaded in the DSC chamber. A sealed empty pan was used as a reference. Samples were scanned at the heating rate of 5 ◦C/min through the temperature range of 50–400 ◦C.

#### 2.4.5. Dynamic Light Scattering (DLS)

The size distributions of the microgel samples subjected to the tested parameters and swelling test at different pH values of 2, 4, 7, and 10 were analyzed by dynamic light scattering (DLS Shimadzu SALD-7000, Tokyo, Japan). pH values in the swelling test were adjusted by using 0.1 M HCl and 0.1 M NaOH.

2.4.6. Optical Microscopy, Scanning Electron Microscopy (SEM), and Transmission Electron Microscopy (TEM)

The optical microscopic morphologies of the reverse micelles in the emulsion and the microgels were visualized using an optical microscope (Olympus CKX41 Inverted Phase Contrast Microscope, Tokyo, Japan) at the magnification of 20×. Morphologies of the freeze-dried microgels were studied by scanning electron microscopy (Desktop SEM Hitachi TM3030 Plus, Tokyo, Japan) and transmission electron microscopy (TEM Hitachi HT7700, Tokyo, Japan). The microgels were freeze-dried by immersing in liquid N2 for 1 h before immediately loading the frozen samples into a chamber of a freeze dryer (Eyela Freeze Dryer FDU-1200, Tokyo, Japan). The freeze-drying process was operated at a condenser temperature of −40 ◦C under high vacuum. For SEM, the freeze-dried microgels were coated with gold using a gold sputter (Quick cool coater SC-701MC, Tokyo, Japan) under a high-vacuum condition. The surface morphology of the coated microgels was then observed at a voltage of 15 kV using a back-scatter detector (BSE) mode at 2000×. For TEM, the freeze-dried microgels were stained with Osmium tetroxide for 20 s before observing the sample morphology at 100 kV and 4000×.

#### 2.4.7. Zeta-Potential

Zeta-potential of the microgels was determined using a zeta-potential analyzer equipped with auto-titrator, stirrer, and inbuilt peristatic pump (Otsuka ELSZ, Tokyo, Japan). The zeta-potential was recorded at the pH values ranging from 2 to 10 adjusted using 0.1 M HCl and 0.1 M NaOH. All measurements were carried out at room temperature (20 ± 5 ◦C).

#### 2.4.8. Brunauer-Emmett-Teller (BET)

N2 adsorption-desorption isotherms of freeze-dried chitin microgels were carried out using a surface area and porosity analyzer (Micromeritics TriStar II, Norcross, GA, USA) at 77 K using Brunauer-Emmett-Teller (BET) and Barrett-Joyner-Halenda (BJH) analyses. Before analysis, samples were degassed at 30 ◦C on a vacuum line for 24 h.

#### **3. Results and Discussion**

#### *3.1. Properties of the Extracted Chitin*

Chitin extraction, in this work, included acid-base hydrolysis and decoloration processes. The crab shell waste was extracted for chitin having yield of 25 ± 8% dry weight.

From XRF analysis, see Table 2, the extracted chitin retained high contents of C and O of the organic compound. The other sea contaminants in the extracted chitin were mainly removed from the crab shells and the quality was similar to the commercial chitin. While the heavy metals in the extracted chitin were not detected as compared to chitin from red shrimp shell [23].

**Table 2.** Elemental composition of crab shell, extracted chitin, commercial chitin, and commercial chitosan analyzed by X-ray fluorescence spectroscopy (XRF).


Figure 2 shows XRD patterns of the extracted chitin obtained in the 2*θ* range of 5–40◦. The diffraction peaks of the extracted chitin, see Figure 2a, and the commercial chitin, see Figure 2b, at 9.4◦, 12.8◦, 19.4◦, 20.8◦, 23.5◦, and 26.4◦ were observed with indices of (020), (101), (110), (120), (130), and (013). These parameters define the crystallographic planes of α-chitin. This indicated that chitin has high molecular packing with inter- or intramolecular hydrogen bonds, imparting a high degree of crystallinity [23–25]. The intensities of the (020) and (110) planes decreased and moved to higher angles with a reduction in the degree of acetylation (DA) [21]. In this work, characteristic peaks of chitosan indexed as (020) and (110) appear at 10.4 and 20.3◦, respectively, see Figure 2d. The extracted chitin exhibited a crystallinity value of 70.3%, as shown in Table 3. The DA of the extracted chitin obtained by XRD and confirmed by FTIR techniques were 55.3% and 60.9%, respectively, see Table 3. This meant that the extracted chitin was partially deacetylated.

**Figure 2.** X-ray diffractograms of crab shell (**a**), extracted chitin (**b**), commercial chitin (**c**), and commercial chitosan (**d**).

**Table 3.** Crystallinity index (%CrI), degree of acetylation (%DA), and peak temperature of extracted chitin, commercial chitin, and commercial chitosan.


As seen in Figure 3, the FTIR spectrum of the extracted chitin had a broad peak at about 3450 cm−<sup>1</sup> assigned to OH stretching. Amide I, II, and III appeared at the observed absorption bands around 1652, 1557, and 1310 cm<sup>−</sup>1, respectively. It was observed that the amide I band of the extracted chitin is split into two 1652 and 1623 cm<sup>−</sup>1. The existence of these interchain bonds of carbonyl groups of amide I and II are responsible for the high chemical stability of the α-chitin structure [23,26]. DSC thermograms of the crab shell, extracted chitin, commercial chitin, and chitosan were compared, see Figure 4. The wide and weak endothermic peak of the extracted chitin in Figure 4b was noticed at about 50–90 ◦C and ascribed to the loss of bound water. The exothermic peak of the crab shell and extracted chitin was observed at 330 ◦C due to the crystalline α-chitin structure. This indicated that the extraction process of chitin retained the α-structure of the resulting product. The extracted chitin had a higher temperature at which the exothermic peak appeared than the chitosan, see Figure 4d. The exothermic peak observed for chitosan at 295 ◦C is the characteristic peak of amine (GlcN) unit decomposition [27].

**Figure 3.** Fourier transform infrared (FTIR) spectra of crab shell (**a**), extracted chitin (**b**), commercial chitin (**c**), and commercial chitosan (**d**).

**Figure 4.** Differential scanning calorimetry (DSC) thermogram of crab shell (**a**), extracted chitin (**b**), commercial chitin (**c**), and commercial chitosan (**d**).

#### *3.2. Reverse Micelle Emulsification for the Fabrication of Chitin Microgels*

#### 3.2.1. Effect of Water, Oil Phase, and Surfactant

In the reverse micelle emulsification, chitin in the alkali solution was prepared at 1, 2, and 3 % *w*/*w*, and then added dropwise into the oil phase. The microgels produced from 1 and 2 % *w*/*w* of the chitin solutions became small in size, but rather aggregated, see Figure 5a,b.

Increasing the chitin concentration to 3% provided more dispersed microgels with an average size of 7.1 ± 0.3 μm, see Figure 5c,g. Nevertheless, microgels produced from 3% chitin appeared as a weak gel with a less uniform size, as seen in Figure 5d,e. These differences are due to the low ratios of oil and water phases (O:W) at 3:1 and 7:1. At O:W of 15:1, the microgel appeared to be more dispersed, see Figure 5f. From the dynamic light scattering analysis, chitin microgels prepared from 1–3% *w*/*w* of chitin and O:W of 15:1 yielded a narrower size distribution (5–10 μm), see Figure 5g. However, there were wider size distributions (10–100 μm) of microgels when O:W of 3:1 and 7:1 were used (Figure 5h). Due to the low O:W of 3:1 and 7:1, the reverse micelles of chitin could not properly disperse in the oil phase during agitation. This might be due to the bigger microgels yielded after gelation, meaning that, low volume of the dispersion phase caused a high incidence of micelle breaking collisions during agitation.

**Figure 5.** Optical microscopic images of microgels prepared from chitin solution at the concentrations of 1% *w*/*w* (**a**), 2% *w*/*w* (**b**), and 3% *w*/*w* (**c**), and O:W of 3:1 (**d**), 7:1 (**e**), and 15:1 (**f**) by controlling the concentration of Span 80 (HLB 4.3) at 5% *w*/*w* in oil phase. Gelation was carried out using 800 μL of 1.0 M HCl. The representative size distributions of microgels prepared by different chitin concentrations and O:W ratios are shown in (**g**) and (**h**), respectively.

It can be clearly observed in Figure 6a,g that HLB 4.3 was suitable for preparing chitin microgels in this study as compared to mixed surfactants having HLB 5 and 6 due to the balance of the size and strength of hydrophilic and lipophilic moieties of surfactant molecules. The bigger microgels with wider size distribution were the result of using mixed surfactants at HLB of 5 and 6, see Figure 6b,c,g. This was due to the higher hydrophilic portion in the surfactant that allowed the chitin aqueous solution to form larger and stable cores inside the reverse micelles.

Span 80 concentrations (HLB 4.3) of 3, 5, and 7 % *w*/*w* were varied in the preparation of chitin microgels. In this range of surfactant concentrations, the morphology of the resulting microgels observed through the optical microscope were similar in size, see Figure 6d–f. The size distributions were also comparable, when the Span 80 concentrations were in the range of 3–7% *w*/*w*, see Figure 6h. However, as seen in Figure 6, the microgel prepared under the condition of 3% *w*/*w* surfactant likely exhibited aggregation, in which slightly a larger portion of ~20 μm microgel was observed.

**Figure 6.** Optical microscopic images of microgels prepared from hydrophilic-lipophilic balance (HLB) values of surfactant at 4.3 (**a**), 5 (**b**), and 6 (**c**), and Span 80 at the concentrations of 3 % *w*/*w* (**d**), 5 % *w*/*w* (**e**), and 7 % *w*/*w* (**f**) by controlling the concentration of the chitin solution at 3 % *w*/*w* and an O:W ratio at 15:1. Gelation was carried out using 800 μL of 1.0 M HCl. The representative size distributions of microgels prepared by different HLB values and Span 80 concentrations are shown in (**g**) and (**h**), respectively.

#### 3.2.2. Effect of Gelation

The gelation was implemented while the alkali chitin solution inside the emulsified reverse micelles was surrounded by an oil phase. When HCl solution was used as the gelling agent, the alkali conditions of the chitin solution have a neutralizing acid-base reaction. From Figure 7a–c,g, the concentration of HCl greatly affected the size distribution. When the HCl concentration was changed to 0.05, 0.1, and 1 M, a size variation (~5, 20, and 40 μm) was observed as the diluted HCl concentration was 0.05 M. It was seen that the size was less variant with increased HCl concentrations. As seen in Figure 7g, concentrations of 0.1 M and 1.0 M provided 50 and 5 μm diameter microgel spheres, respectively. This was related to the effect of the water phase on the microgel preparation, in which the initial concentration of chitin solution and external water affected the size distribution of the microgels. It was possibly owing to the involvement of a higher amount of water in the emulsion system; Span 80 was able to hold/absorb water into the core of reverse micelles [14]. The effect of a volume of 1.0 M HCl was further studied, see Figure 7d–f,h. It was found that only 400 μL was sufficient to yield the narrow size distribution of ~5 μm microgels. However, increasing the volume to 800–1200 μL slightly increased the portion of ~5 μm microgels.

**Figure 7.** Optical microscopic images of microgels prepared from gelation using HCl at the concentrations of 0.05 M (**a**), 0.1 M (**b**), and 1.0 M (**c**), and 400 μL (**d**), 800 μL (**e**), and 1200 μL (**f**) of 1.0 M HCl by controlling the concentration of chitin solution of 3% *w*/*w* at an O:W ratio of 15:1, and concentration of Span 80 (HLB 4.3) at 5% *w*/*w* in the oil phase. The representative size distributions of microgels prepared by different concentrations and volume of HCl are shown in (**g**) and (**h**), respectively.

#### *3.3. Properties of Chitin Microgels Prepared by the Reverse Micellar Method*

The appearance of chitin solution containing reverse micelles is shown in Figure 8a, corresponding to microgels prepared with 3% *w*/*w* chitin solution, O:W of 15:1, 5% *w*/*w* Span 80 (HLB 4.3), and gelation by 800 μL of 1 M HCl. As compared to the size of reverse micelles, no remarkable change in size was observed after gelation of chitin microgels. Figure 8b showed no microgel breakage induced by collision during stirring. Electron microscopic images of the freeze-dried samples revealed a spherical chitin microgel with macropores on the surface, see Figure 8c, and an internal porous structure, see Figure 8d. The formation mechanism of macropores on microspheres prepared by an emulsion system when using Span 80 for the Poly(styrene-divinyl benzene) system has been reported [14]. Similarly, macropores of the resultant chitin microgels are strongly related to the absorption of water from the external aqueous phase into the reverse micelles. Since Span 80 with a HLB of 4.3 was less hydrophobic, it has a stronger ability to absorb water.

**Figure 8.** Optical microscopic images of reverse micelles in emulsion (**a**) and microgels (**b**) prepared under 3 % *w*/*w* chitin solution, O:W = 15:1, 5% *w*/*w* Span 80 (HLB 4.3), and gelation by 800 μL of 1.0 M HCl. Morphologies of the freeze-dried microgels are shown at magnifications of 2000× by scanning electron microscopy (SEM) (**c**) and 4000× by transmission electron microscopy (TEM) (**d**).

The porosity data of the chitin microgels was characterized by BET. The BET isotherms of the microgels are shown in Figure 9. The chitin microgels showed a type II isotherm for a microporous material according to the IUPAC classification [28]. The surface area and pore volume were 22.6 m2/g and 0.03 cm3/g, respectively.

**Figure 9.** Brunauer-Emmett-Teller (BET) isotherm of freeze-dried microgels prepared under conditions of 3 % *w*/*w* chitin solution, O:W = 15:1, 5% *w*/*w* Span 80 (HLB 4.3), and gelation by 800 μl of 1.0 M HCl.

Under the optimal conditions of microgel preparation, the charge of partially deacetylated chitin microgels was investigated by measuring the zeta potential according to various pHs. Figure 10 shows the zeta potential of the chitin microgels prepared at 3% *w*/*w* chitin solution, O:W of 15:1, 5% *w*/*w* Span 80 (HLB 4.3), and gelation by 800 μL of 1.0 M HCl. The positive zeta potential at pH below 6 clearly indicated partially deacetylated chitin with different %DA due to the different number of amine groups to be protonated, leading to the positive charges. The isoelectric point (IP), that is the null zeta potential, increases to a higher pH value with lower %DA. Accordingly, IP values of chitosan (DA < 50%) and chitin (DA > 50%) were detected at pH values of 8.2–8.8 and 7.3–7.6, respectively [24,29]. In the present work, the IP value of the chitin microgels was observed at about pH 7.6, see Figure 10. This confirms that the chitin microgels were not further deacetylated during the preparation process of the chitin microgels.

**Figure 10.** Zeta potential of chitin microgels prepared under conditions of 3 % *w*/*w* chitin solution, O:W = 15:1, 5% *w*/*w* Span 80 (HLB 4.3), and gelation by using 800 μL of 1.0 M HCl.

Since the microgels presented pH dependence, it showed different swelling of the microgels tested over the entire range of pH from 10 to 2. From Figure 11a, the chitin microgels swelled at low pH (pH < IP). Microgels that were approximately 6 μm in size increased to ~60 μm. However, the gradual decrease was observed between pH 2 and pH 4 due to the similar zeta potential of approximately +25 mV, as seen in Figure 10. This was possibly the impact of the degree of acetylation, which controls the characteristics and activities of chitin [30]. In the extracted chitin with a moderate degree of acetylation (~60%), a number of amino groups in the chitin polymer chains were protonated while exposed to a specific pH. The protonation leads to the repulsion of polymer chains and allows more water to enter into the microgel network; consequently, swelling occurs [31]. After adjusting the pH backward, from 2 to 10, a reversible swelling-shrinking behavior was noticed. The chitin microgels began to shrink to a smaller size as the pH increased (pH > IP), see Figure 11b. This was because the deprotonation made the electrostatic interactions in the microgel network reconstruct [31].

**Figure 11.** Reversible swelling-shrinking behavior of chitin microgels prepared under conditions of 3 % *w*/*w* chitin solution, O:W = 15:1, 5% *w*/*w* Span 80 (HLB 4.3), and gelation by using 800 μL of 1.0 M HCl. Responsiveness to pH was tested from pH 10 towards pH 2 (**a**) and pH 2 towards pH 10 (**b**).

#### **4. Conclusions**

Chitin extracted from crab shell waste was used for microgel fabrication. Simple gelation inside the emulsified reverse-micelles with low energy consumption was applied to prepare the chitin microgels. The spherical size distribution and the morphology of the microgels were greatly affected by the volume of the dispersion phase, hydrophilic-lipophilic balance of the used Span 80 surfactant, and concentration of the gelation agent. As a result, the chitin microgel with narrow size distribution (average size of 7.1 ± 0.3 μm) and porous spherical morphology was achieved under the condition of 3 % *w*/*w* chitin solution, O:W of 15:1, 5% *w*/*w* Span 80 (HLB 4.3), and gelation by 800 μL of 1.0 M HCl. Moreover, the prepared chitin microgels exhibited pH-dependent swelling-shrinking behavior over a wide range of pH values of between 2–10.

**Author Contributions:** Conceptualization, S.T.; methodology, M.O.; validation, S.T. and T.K.; formal analysis, M.O.; investigation, M.O. and S.T.; resources, S.T. and T.K.; writing—original draft preparation, S.T.; writing—review and editing, S.T.; supervision, S.T. and T.K.; project administration, S.T.

**Funding:** This research was funded by Nagaoka University of Technology.

**Acknowledgments:** This work was supported by Nagaoka University of Technology. The authors thank Analysis and Instrumentation Center of Nagaoka University of Technology for the technical assistance.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **Chitosan Nanoparticles Rescue Rotenone-Mediated Cell Death**

#### **Jyoti Ahlawat 1, Eva M. Deemer <sup>2</sup> and Mahesh Narayan 1,\***


Received: 13 March 2019; Accepted: 4 April 2019; Published: 11 April 2019

**Abstract:** The aim of the present investigation was to study the anti-oxidant effect of chitosan nanoparticles on a human SH-SY5Y neuroblastoma cell line using a rotenone model to generate reactive oxygen species. Chitosan nanoparticles were synthesized using an ionotropic gelation method. The obtained nanoparticles were characterized using various analytical techniques such as Dynamic Light Scattering, Scanning Electron Microscopy, Transmission Electron Microscopy, Fourier Transmission Infrared spectroscopy and Atomic Force Microscopy. Incubation of SH-SY5Y cells with 50 μM rotenone resulted in 35–50% cell death within 24 h of incubation time. Annexin V/Propidium iodide dual staining verified that the majority of neuronal cell death occurred via the apoptotic pathway. The incubation of cells with chitosan nanoparticles reduced rotenone-initiated cytotoxicity and apoptotic cell death. Given that rotenone insult to cells causes oxidative stress, our results suggest that Chitosan nanoparticles have antioxidant and anti-apoptotic properties. Chitosan can not only serve as a novel therapeutic drug in the near future but also as a carrier for combo-therapy.

**Keywords:** chitosan (CS); anti-oxidant; anti-apoptotic activity; rotenone; Parkinson's disease (PD)

#### **1. Introduction**

Parkinson's disease (PD) is a multifocal progressive neurodegenerative disorder clinically defined by the presence of akinesia, postural instability, muscular rigidity, and tremor [1]. It is the second most common neurodegenerative disease and is prevalent in 0.1–0.3% population with an increased frequency observed in patients ≥65 years [2]. Interestingly, PD patients often display non-motor signs and symptoms such as sleep disturbances, mood deflection, anosmia, gastrointestinal dysfunction (e.g., 80% of patients suffer from constipation), sexual-urinary dysfunction, thermoregulation changes, neuropsychiatric problems, cardiovascular disturbances, and pain [1,3]. PD is characterized by the selective loss of dopaminergic neurons in the substantia nigra pars compacta and arises due to the deposition of insoluble polymers of α-synuclein in the neurons, forming spherical intracytoplasmic inclusions known as Lewy bodies [1]. These lamellated cytoplasmic bodies eventually result in neurodegeneration and the death of dopaminergic neurons [1]. This neuronal death is associated with disruption in cellular hemostasis, resulting in disruption of the nuclear membrane integrity, signaling α-synuclein aggregation which later propagates to other neurons by direct or indirect means [4]. Furthermore, studies have shown that α-synuclein aggregation impairs axonal transport and exerts a detrimental effect on the health of neurons due to the activation of neighboring inflammatory microglial cells [3].

Parkinson's disease can be either inherited or sporadic in nature. Although, familial PD accounts for around 10% of the cases, sporadic PD has been found to account for the remaining ones. Moreover, the etiology of PD is not completely understood but sporadic PD is believed to originate from

interaction of individual genetic susceptibility and environmental exposure [5,6]. Probably, there is not one single factor that is solely responsible for causing the disease. Rather, there exists several factors acting simultaneously [7]. Previous studies have suggested that pesticides such as rotenone are involved in the increased risk of Parkinson's disease [5]. In addition to being a pesticide, it is a potent, highly specific inhibitor of complex I of the mitochondrial electron transport chain (ETC) [6]. Unlike N-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP), which causes defects in complex I of ETC in catecholaminergic neurons, rotenone causes complex I inhibition, uniformly, across the brain [8]. Its hydrophobic structure allows easy penetration through the blood–brain barrier and cell membrane. The selective toxicity of this lipophilic compound is relevant because of its wide usage as a herbicide in gardens and as a delousing agent for animals and humans [7]. Furthermore, studies have shown degradation of selective nigrostriatal dopaminergic neurons upon rotenone infusion, reproducing pathological features of clinical Parkinson's disease [6]. Moreover, a study by Niyanyu et al. showed that rotenone can induce mitochondrial reactive oxygen species (ROS) production which is closely related to rotenone-induced apoptosis [9].

Since no viable treatment exists for PD, there is an urgent and unmet need for the development of novel therapeutic agents to either cease or reverse the symptoms or the progression of this progressive age-related disorder [10]. Synthetic compounds are associated with various side-effects [11]. Therefore, there is a need to find some natural neuroprotective agent that has the ability to scavenge ROS and hence defer the progression of Parkinson's disease [10]. Chitosan is a cationic polysaccharide, composed of a linear chain of D-glucosamine and N-acetyl-D-glucosamine linked via a β (1,4) bond, obtained from an alkaline N-deacetylation of chitin [12]. This marine shrimp-derived carbohydrate possesses well-documented antioxidant properties with minimal or no side-effects observed. Moreover, they also exhibit neuroprotective, anti-hemorrhagic, anti-tumor, anti-diabetic, anti-viral, and antibacterial effects. Furthermore, it has mucoadhesive properties allowing easy penetration of this carbohydrate through the well-organized epithelia [13]. In a study, in 2001, Gilgun-Sherki et al. showed that elevated ROS production and an imbalance between pro-oxidant and antioxidant activity (e.g., superoxide dismutase, catalase, and glutathione peroxidase enzyme) leads to neuronal death and hence a diseased condition [14]. In a different study, Guo et al. in 2006 reported that fucoidan (sulfated polysaccharide) could reverse changes such as superoxide dismutase activity and alleviate the reactive oxygen species level in PC 12 cells when exposed to hydrogen peroxide [15]. Later, Gao et al. (2012) showed the antioxidant effect of fucoidan on hydrogen-peroxide-treated PC12 cells and the pathway associated with it [16]. In a different study, Xie et al. (2014) reported antioxidants could alleviate the reactive oxygen species level [17]. Further, in 2016, Wang et al. reported fucoidan pretreatment could rescue the cells from oxidative stress, protein carbonyl lipid peroxidation, and mitochondrial dysfunction [18,19]. Related to this, Liu et al. (2016) showed the effect of sulfonated chitosan on the differentiation of neuronal cells and exhibited immunomodulatory effects [20]. Recently, Magnigandan et al. (2018) reported the anti-oxidant and ROS scavenging activity of low molecular weight sulphonated chitosan, where they found that rotenone insult resulted in antioxidant depletion and lipid oxidation causing cellular damage, oxidative stress, mitochondrial dysfunction, and hence, a diseased state which were reversed by the action of low molecular weight sulfated chitosan [11].

However, there are no reports on the in vitro neuroprotective effects of bare chitosan nanoparticles. Therefore, the goal of this study was to exploit the antioxidant and anti-apoptotic activity of the prepared chitosan nanoparticles for evaluation in vitro, against a human SH-SY5Y neuroblastoma cell line. Hence, rotenone was used as the causative agent for inducing PD in the SH-SY5Y cell line and then, the therapeutic neuroprotective efficacy of the synthesized chitosan nanoparticles was evaluated. Therefore, we propose that chitosan nanoparticles might be a potential candidate for the prevention of PD.

#### **2. Materials and Methods**

#### *2.1. Materials*

Chitosan (>75% deacetylated), and Sodium tripolyphosphate (Na-TPP) were purchased from Sigma-Aldrich (Saint Louis, MO, USA). Whereas, Dimethyl sulfoxide and acetic acid were ordered from Fisher Chemical (Hampton, NH, USA).

#### *2.2. Preparation of Chitosan Nanoparticles Using Ionotropic Gelation Method*

Chitosan nanoparticles were synthesized using the Calvo et al. 1997 method [21]. In this, 0.175% (w/v) chitosan powder was dissolved in 1% (v/v) acetic acid and kept on a magnetic stirrer for overnight stirring at a temperature between 25–28 ◦C. Later, the pH of the solution was adjusted to 5.2 using 1M NaOH followed by addition of 0.1% (w/v) sodium tripolyphosphate in a dropwise fashion. The chitosan solution was then stirred at 1000 rpm for 10 min. The solution was centrifuged at 20,000 rpm (Sorvall RC-5B refrigerated centrifuge, Fisher Scientific, Hampton, NH, USA) for 90 min at a temperature of 4 ◦C to pelletize the chitosan nanoparticles. After centrifugation, the supernatant was discarded and the pellet was washed with deionized water three times. Ultrasonication was performed using a probe sonicator (Branson Sonifier 450, Emerson Electric Company, St. Louis, MO, USA) in an ice bath for 10 min at an amplitude of 30%. The obtained suspension was then immediately freeze dried using lyophilizer (Labconco, Kansas City, MO, USA). After freeze-drying, the samples were stored at 4 ◦C to carry out further analysis.

#### *2.3. Characterization of Nanoparticles*

#### 2.3.1. Determination of Average Size, Polydispersity Index, and Zeta Potential Using Dynamic Light Scattering

The average size, polydispersity index, and surface charge of the nanoparticles was determined using DLS (Dynamic light scattering) at 25 ◦C. A total of 1 mL CS NP solution was diluted ten times in deionized water and the sample was analyzed using a Malvern Zetasizer ZS90 (Malvern Panalytical Ltd., Malvern, UK).

#### 2.3.2. Scanning Electron Microscopy (SEM)

The surface morphology of the prepared chitosan nanoparticles was analyzed using the S-3400N Type II scanning electron microscope, Pleasanton, USA (Hitachi High-Technologies Corporation, Tokyo, Japan). The liquid samples on a glass slide were dried overnight in a sterilized fume hood and later, the glass slide was mounted on a stainless stub using double-sided carbon tape. The sample was then coated with gold using a gold sputterer for 30 s to make the sample conductive. The images were recorded at an accelerating voltage of 2 kV and probe current below 20 μA.

#### 2.3.3. Transmission Electron Microscopy (TEM)

The morphology of the nanoparticles was studied using the H-7650 transmission electron microscope, manufactured by Hitachi High-Technologies Corporation in Pleasanton, CA, USA. A single drop of the CS Nanoparticle dispersion was placed on carbon 400 mesh copper grid (Ted Pella, Redding, CA, USA). The copper grids were then dried overnight in a sterile fume hood and the images were taken using the Quartz PCI version 8 software (Quartz Imaging Company, Vancouver, BC, Canada) in TEM mode (200 kV).

#### 2.3.4. Atomic Force Microscopy (AFM)

The surface roughness and morphology of the chitosan nanoparticles was further investigated using AFM (NT-MDT NTEGRA) in non-contact mode using Ted Pella TAP 150AL-G tip (Redding, CA, USA) with a radius of <10 nm. After capturing the images, the data analysis was performed using NOVA software (NOVA Company, ChongQing, China).

#### 2.3.5. Fourier Transform Infrared Spectroscopy (FTIR)

The IR spectra of the Chitosan nanoparticle sample was obtained using a Nicolet, Thermo Scientific FTIR instrument (Waltham, MA USA). The sample was ground to fine powder along with a KBr pellet. The scanning range was from 500–4000 cm−1. The data were analyzed using OMNIC software (Fisher Scientific, Hampton, NH, USA).

#### *2.4. Cellular Behavior of Human SH-SY5Y Neuroblastoma Cell Lines on Treatment with Samples*

#### 2.4.1. Cell Culture

Annexin V- FITC apoptosis kit (Beckman Coutler, Brea, CA, USA), Hoechst 33342 fluorescent satin (Invitrogen, Carlsbad, CA, USA), propidium iodide (PI) (Invitrogen, Eugene, OR, USA), Fetal Bovine Serum (FBS) (Atlanta Biologicals, Atlanta, GA, USA), and a human neuroblastoma cell line SH-SY5Y (ATCC, Manassas, VA, USA) were purchased. SH-SY5Y cells were cultured in DMEM and Hans's F12 media mixture (1:1) comprising of 10% FBS (v/v) supplemented with 1% (v/v) penicillin-streptomycin and maintained at 37 ◦C in an incubator with 5% CO2 atmosphere. Cells were sub-cultured every 48 h and Trypsin-EDTA 0.25% (1×) was used to detach cells from the culture surface when needed.

#### 2.4.2. Differential Nuclear Staining Cytotoxicity Assay

The cytotoxicity of different concentrations of chitosan nanoparticles, chitosan powder, and rotenone were evaluated. Cells were first cultured on 96-well plates and incubated for 24 h to allow attachment to the culture surface. Later, cells were treated with a different concentration of rotenone and chitosan nanoparticles to determine the possible cytotoxic effect of the added treatments. Subsequently, untreated cells were taken as a negative control and hydrogen-peroxide-treated cells were taken as a positive control. Moreover, to determine the cytotoxic effect of rotenone on the cells, the cells were treated with chitosan nanoparticles (different concentrations) 6 h prior to rotenone exposure. Subsequently, cells were further incubated for 24 h. Later, 1 μg/mL mixture of PI/Hoechst 33342 was added to each well in the 96-well plates 1 h prior to the imaging process [22]. The images were captured using a Bioimager system (BD Biosciences Rockville, Montgomery, MD, USA). Five images were taken per well using a 10× objective lens. Subsequently, BD AttoVision v1.6.2 software (BD Biosciences Rockville, Montgomery, MD, USA) was used to determine the percentage cell death per well.

#### 2.4.3. Flow Cytometric Assay

The SH-SY5Y cell lines were seeded at a density of 20,000 cells per well in a 24-well plate and incubated for 24 h to allow attachment of the cells to the culture surface. Cells were then incubated with various concentrations of chitosan nanoparticles prior to rotenone exposure and subsequently, cells were incubated for an additional 24 h. Cells from each well were then collected, washed, and processed [23]. Briefly, cells were concurrently stained by suspending them in a solution containing annexin V-FITC (PI) in 100 μL of binding buffer (Beckman Coulter, Brea, CA, USA). After incubation for a time interval of 15 min on ice, in a sterilized Lab Safety Cabinet II in dark, 400 μL of ice-cold binding buffer was added to the cells. The resulting suspension was then homogenized gently and subsequently, analyzed using a Cytomics FC 500 Beckman Coulter Flow cytometer. For each sample, approximately 10,000 events were captured and data analysis was performed using Beckman Coulter CXP software.

#### 2.4.4. Mitochondrial Membrane Potential (ΔΨm)

The loss of mitochondrial membrane potential is a hallmark for cellular apoptosis. Briefly, SH-SY5Y cells were seeded onto a 96-well plate for 12 h. Subsequently, cells were treated with chitosan

nanoparticles 6h prior to rotenone treatment. Later, after 24 h of incubation, the cells were incubated with rhodamine 123 dye at 37 ◦C for 30 min. Finally, the mitochondrial membrane potential was evaluated quantitatively using a Bioimager system (BD Biosciences Rockville, Montgomery, MD, USA) and the fluorescence intensity was measured at 485/530 nm.

#### *2.5. Statistical Significance*

The experimental data were expressed as the mean ± standard deviation of one or more individual experiments wherever applicable. The analysis of experimental data was performed with the students t-test using Graph Pad Prism 6.0 (San Diego, CA, USA) and statistically significant values were indicated as *p* < 0.05.

#### **3. Results and Discussion**

#### *3.1. Dynamic Light Scattering*

Particle size and surface charge are two important factors determining the size, stability, and effective delivery of the drug to the target site [24]. Figure 1A depicts the average particle size distribution of the synthesized CS Np using an ionotropic gelation method. As can be observed, the bare nanoparticles have an average size of 197.8 ± 49.18 nm with a PDI 0.244.

**Figure 1.** Dynamic light scattering (DLS) images of (**A**) average particle size distribution of CS Nanoparticle and (**B**) Zeta potential of synthesized CS Nanoparticle.

The zeta potential is a measure of the stability and surface charge on the particles. A high zeta potential is indicative of high electric surface charge allowing strong repulsion between the surrounding particles and hence preventing aggregation [24]. Figure 1B depicts the average zeta potential value of CS Np which was observed to be +36.0 ± 4.68. A previous study on chitosan nanoparticles showed that the blank nanoparticles in the ratio 5:1 CS/TPP had a particle size in the range of 300 to 390 nm and a zeta potential of + 44 ± 5.2, which strongly supports our findings [25].

#### *3.2. Morphological Characterization*

Figure 2 displays the scanning electron micrograph images of chitosan powder, freshly prepared nanoparticles, and lyophilized chitosan nanoparticles sample. It was observed that the nanoparticles displayed spherically compact structures with an average diameter of 220 ± 40 nm, see Figure 2E, which almost coincides with our dynamic light scattering study. Similar results were reported in another study where the CS/TPP in a 5:1 ratio displayed an average size of 200 ± 24 nm [26,27]. The freshly prepared CS Np appeared as clusters. This can be attributed to the fusion of particles through hydrogen bonding [24]. Moreover, the stability of chitosan nanoparticles was tested in buffer solution (pH 7.4). The morphology of the nanoparticles was observed to be spherical with an average diameter of 200 nm, see Supplementary Figure S1.

**Figure 2.** Scanning Electron Microscopy images of (**A**) chitosan powder; (**B**) freshly prepared CS Np and (**C**,**D**) lyophilized CS Np. Further, (**E)** depicts the average size distribution of the nanoparticles obtained from scanning electron microscopy. Whereas, (**F**) depicts the transmission electron microscopic image of chitosan nanoparticles.

Transmission electron microscopy is a technique which provides information on the morphology and size of the particles. Figure 3 depicts a TEM micrograph of small and spherical chitosan nanoparticles with a diameter of around 120 ± 30 nm. The size of the nanoparticles in the TEM image, see Figure 2F, are smaller than that represented by SEM and DLS. This can be attributed to the aggregation of nanoparticles due to their high surface area and energy which generates a larger entity [28]. However, dimensions of freshly prepared single particles can be observed clearly in the TEM image of the CS Np sample.

**Figure 3.** (**A**) 2D and (**B**) 3D atomic force micrograph of chitosan nanoparticles in a 2-micron × 2-micron area.

#### *3.3. AFM Analysis*

Scanning electron microscopy and transmission electron microscopy provides a two-dimensional projection or image of the nanoparticles. However, atomic force microscopy is a powerful technique which enables a three-dimensional surface profile of the nanoparticles to be viewed. In addition, it can also provide accurate heights of the nanoparticles. Figure 3 shows the atomic force micrograph of chitosan nanoparticles. As can be observed, the nanoparticles appeared to be spherical in shape. Furthermore, the size distribution histogram was performed for the nano-chitosan in a 2-micron × 2-micron area. Analysis shows that the average particle size is around 200 nm as the histogram peaks are at 0.2 microns, see Supplementary Figure S2, which coincides with our DLS, SEM, and TEM data. Similar results were observed in previous reports [29–31]. In addition, a small peak appeared at 600 nm which can be attributed to the fusion of particles through hydrogen bonding [24].

#### *3.4. Spectroscopic Characterization*

Fourier-transform infrared spectroscopy (FTIR) is a powerful spectroscopic technique to determine the chemical composition and presence of a drug inside the nanoparticles. Figure 4 represents the FTIR spectra of chitosan powder and chitosan nanoparticles. In the FTIR spectrum of chitosan nanoparticles, a broad peak at 3436 cm−<sup>1</sup> corresponds to the -NH and -OH stretching vibrations. The weak band at 2930 cm−<sup>1</sup> corresponds to –CH stretching, whereas vibrational bands at 1640 cm−1, 1560 cm−1, and 1320 cm−<sup>1</sup> may be attributed to the amide carbonyl stretch and -NH bend of the amine groups in the chitosan nanoparticles [32]. The band at 1099 cm−<sup>1</sup> represents C–O bond stretching. Moreover, the vibrational band at 860 cm−<sup>1</sup> corresponds to the CH2OH group in the pyranose ring of chitosan. The only difference between the spectra of chitosan powder and chitosan nanoparticles occurred at 1155 cm−1, which could be assigned to the linkage between phosphate groups of the sodium tripolyphosphate and ammonium ions of the chitosan group. Furthermore, our results agree with report of Gopalakrishnan et al. (2014) and Jafary et al. (2016) [24,32].

**Figure 4.** Fourier-transform infrared spectroscopy (FTIR) spectra of chitosan nanoparticles and chitosan powder.

#### *3.5. Effect of Chitosan Nanoparticles on Rotenone-Induced Cell Death*

The cytotoxicity of chitosan nanoparticles and its protective function against rotenone insult in SH-SY5Y cell line were checked using a high-throughput screening assay. Figure 5i shows cells exposed to an increasing concentration (1–20 μM) of chitosan nanoparticles exhibited cytotoxicity from 9% to 21%. A 10 μM concentration did not display much difference compared to the untreated and vehicle controls. However, on addition of 50 μM rotenone, around 35–50% of the cell death was

reported in the SH-SY5Y cell line after 24 h of incubation time, see Figure 5ii. In contrast, the addition of a 10 μM chitosan nanoparticle solution prior to rotenone exposure resulted in 14–20% cell death compared to 35–50% cell death upon rotenone administration. These data show that treatment of cells with chitosan prior to rotenone exposure attenuated cell death by 25–30%. The morphology of cells after rotenone treatment and cells treated with chitosan prior to rotenone exposure further displays the protective aspect of chitosan nanoparticles against rotenone insult. Bright field microscopy images and Hoechst 33342-propidium iodide staining pictures, see Figure 5iii, further support the protective aspect of chitosan nanoparticles. The pretreatment of cells with a 10 μM chitosan nanoparticles solution for 6 h enabled the cells to retain their cellular morphology, as depicted by bright field microscopy images. In contrast, cells treated with 50 μM rotenone showed a morphology similar to that of cells exposed to 50 μM hydrogen peroxide, see Figure 5iv. Hydrogen peroxide was used as a positive control at a concentration of 50 μM.

**Figure 5.** *Cont*.

**Figure 5.** (**i**) Cytotoxicity of chitosan nanoparticles at different concentrations after 24 h of incubation; (**ii**) Protective effect of chitosan nanoparticles against rotenone insult in SH-SY5Y cells; (**iii**) Hoechst-propidium iodide staining images of (**a**) untreated cells, (**b**) water + Tripoly phosphate; (**c**) chitosan nanoparticles 10 μM concentration, (**d**) 50 μM rotenone, (**e**) CS Np + Rotenone and (**f**) hydrogen-peroxide-treated cells after 24 h of treatment; (**iv**) Bright field microscopy images of (**a**) untreated cells, (**b**) CS Np 10 μM, (**c**) CS Np + RT, & (**d**) 50 μM RT cells visualized using a compound microscope after 24 h of treatment. Each experimental point was assessed in quintuplicate.

#### *3.6. Flow Cytometry Analysis*

The mechanism by which cell-death occurs (apoptotic pathway or necrotic pathway) was examined. Figure 6i,ii shows cells pre-treated with 20 μM chitosan powder, 10 μM chitosan nanoparticles, and 20 μM chitosan nanoparticles. A concentration of 50 μM hydrogen peroxide was used as a positive control whereas untreated cells were used as a negative control. As can be observed, cells treated with 50 μM rotenone showed a substantial increase in early and late apoptosis (lower right and top right quadrants of the plot) 24 h after administration. A similar result was observed when cells were treated with 50 μM hydrogen peroxide. However, prior treatment of cells with 20 μM chitosan powder and 10 μM and 20 μM chitosan nanoparticles resulted in a notable rescue of the cells from rotenone insult (apoptotic-cell death). A notable difference was observed when 50 μM rotenone and CS Np 10 μM + rotenone treatments were compared. Thus, we can conclude that chitosan nanoparticles exhibit anti-apoptotic activity.

**Figure 6.** *Cont*.

**Figure 6.** (**i**) Flow cytometry matrix plots used to measure apoptosis/necrotic distribution for (**A**) untreated, (**B**) 20 μM Chitosan powder + RT, (**C**) CS Np 20 μM, (**D**) CS Np+ RT, (**E**) 50 μM RT, and (F) hydrogen peroxide after 24 h of treatment; (**ii**) quantification of apoptotic/necrotic assay under the previously mentioned conditions. Each experimental point was assessed in triplicate.

#### *3.7. Chitosan Np Prevents Rotenone-Induced Mitochondrial Dysfunction*

The untreated cells, see Figure 7(iA), and cells treated with chitosan nanoparticles, see Figure 7(iB), exhibited green fluorescence, indicating that a large fraction of mitochondria inside the cell were in an energized state. However, cells upon treatment with rotenone, see Figure 7(iD), displayed decreased mitochondrial energy transduction as observed by the disappearance of green fluorescence. Further, cells treated with chitosan nanoparticles prior to rotenone treatment, see Figure 7(iC), exhibited green fluorescence which can be attributed to the inhibition of the collapse of the membrane potential via rescue of mitochondrial membrane depolarization by chitosan nanoparticles. Thus, these results suggest that chitosan nanoparticles may inhibit rotenone-induced cellular apoptosis through a mitochondria-involved pathway, as revealed by the considerable increase in green fluorescence intensity.

**Figure 7.** *Cont*.

**Figure 7.** The effect of CS Np on rotenone insult determined using rhodamine 123 dye. (**i**) Cells were incubated with rhodamine 123 and the mitochondrial membrane potential was monitored using a fluorescent microscope for (**A**) untreated, (**B**) CS Np 20 μM, (**C**) 20 μM Chitosan powder + RT, and (**D**) 50 μM RT; (**ii**) quantification of Mitochondrial membrane potential assay under previously described conditions. Each experimental point was assessed in triplicate. The scale bar in the image is 50 μm.

#### **4. Proposed mechanism**

We henceforth propose a mechanistic model for the action of chitosan nanoparticles (Figure 8). Firstly, chitosan nanoparticles are internalized by the cells through endocytosis via an endosome–lysosome pathway. Later, these nanoparticles are released into the cytoplasm from the lysosome due to the low pH environment and enzymatic activity. These nanoparticles then scavenge the reactive oxygen species generated by the inhibition of complex I of the ETC of mitochondria by the rotenone insult. Thus, the apoptotic stimuli exerted on mitochondria is reduced. Subsequently, cytochrome c cannot be released through the permeability transition pore. Further, caspase 3 cannot be activated by caspase 9, resulting in inhibition of Poly ADP ribose polymerase (PARP) cleavage and hence, cell survival [22,23]. Therefore, chitosan nanoparticles protect the cell from rotenone insult due to its anti-oxidant and anti-apoptotic activity.

**Figure 8.** Mechanistic model of the action of chitosan nanoparticles.

#### **5. Conclusions**

The literature is flooded with various number of reports suggesting that oxidative stress and consequent cellular apoptosis are important factors in rotenone-induced cell death. Therefore, pre-treatment of cells with antioxidants such as low molecular weight sulphonated chitosan, fucoidan, quercetin, flavonoids, etc., likely reduces the adverse effects arising from rotenone insult. Although pre-treatment of cells with the above mentioned anti-oxidants were observed to alleviate reactive oxygen species production, the antioxidant effects of chitosan nanoparticulate system against rotenone insult had not been studied previously. In this regard, our study demonstrates the synthesis of chitosan nanoparticles by an ionic gelation method using TPP as the cross-linking agent. The optimized ratio of CS/TPP was 5:1, which produced spherical nanoparticles with an average size of 200 nm as

observed by SEM, TEM, DLS, and AFM. Further, the cytotoxicity of the chitosan nanoparticles and its anti-oxidant and anti-apoptotic effects against rotenone-induced cell death were determined using a differential nuclear staining cytotoxicity assay and flow cytometric analysis. Therefore, Chitosan Nanoparticles could inhibit the rotenone-induced mitochondria involved apoptosis pathway, as revealed by the considerable increase in the green fluorescence intensity in the MMP assay. These findings suggest that chitosan nanoparticles might be a useful and promising neuroprotective agent for the prevention of Parkinson's disease.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/1996-1944/12/7/1176/ s1, Figure S1: SEM image of Chitosan nanoparticles in buffer solution, Figure S2: Size distribution analysis from AFM measurement.

**Author Contributions:** Conceptualization, J.A.; Methodology, J.A.; Software, E.D.; Validation, J.A., and M.N.; Formal Analysis, M.N.; Investigation, M.N.; Resources, M.N.; Data Curation, J.A. and E.D.; Writing-Original Draft Preparation, J.A. and M.N.; Writing-Review & Editing, J.A. and M.N.; Visualization, E.D.; Supervision, M.N.; Project Administration, M.N.; Funding Acquisition, M.N.

**Funding:** This research was funded by NIH grant number 1SC3 GM111200 01A1.

**Acknowledgments:** M.N. and J.A. are thankful to staff of cytometry, screening and imaging core facility of the Border Biomedical Research Center at the University of Texas at El Paso (UTEP). J.A. would also like to thank Ms. Lois Mendez for her help while performing experiments relating to cell studies.

**Conflicts of Interest:** Authors declare that there is no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **Development and Characterization of Bacterial Cellulose Reinforced with Natural Rubber**

#### **Kornkamol Potivara and Muenduen Phisalaphong \***

Department of Chemical Engineering, Faculty of Engineering, Chulalongkorn University, Phayathai Road, Pathumwan, Bangkok 10330, Thailand

**\*** Correspondence: muenduen.p@chula.ac.th; Tel.: +66-2-218-6875

Received: 30 May 2019; Accepted: 17 July 2019; Published: 21 July 2019

**Abstract:** Films of bacterial cellulose (BC) reinforced by natural rubber (NR) with remarkably high mechanical strength were developed by combining the prominent mechanical properties of multilayer BC nanofibrous structural networks and the high elastic hydrocarbon polymer of NR. BC pellicle was immersed in a diluted NR latex (NRL) suspension in the presence of ethanol aqueous solution. Effects of NRL concentrations (0.5%–10% dry rubber content, DRC) and immersion temperatures (30–70 ◦C) on the film characteristics were studied. It was revealed that the combination of nanocellulose fibrous networks and NR polymer provided a synergistic effect on the mechanical properties of NR–BC films. In comparison with BC films, the tensile strength and elongation at break of the NR–BC films were considerably improved ~4-fold. The NR–BC films also exhibited improved water resistance over that of BC films and possessed a high resistance to non-polar solvents such as toluene. NR–BC films were biodegradable and could be degraded completely within 5–6 weeks in soil.

**Keywords:** bacterial cellulose; natural rubber; reinforcing; biodegradable polymers

#### **1. Introduction**

Pollution deriving from plastic materials is becoming one of the most prominent environmental concerns of recent years. Accumulation of plastic products in the environment has harmful effects on wildlife and their environment. Plastics dispersed in ocean ecosystems have become a major pollutant that has led to the direct deaths of marine animals. Therefore, developing renewable and biodegradable materials to replace conventional plastic materials is an increasingly important research area to reduce the level of plastic waste.

Thailand is the largest producer and exporter of natural rubber (NR) globally. Global natural rubber production in 2015 was 12.3 million tons, 92% of which was produced in the Asia-Pacific region. Thailand produced around 4.5 million tons and exported about 3.7 million tons in 2015 [1]. According to a recent release from the Association of Natural Rubber Producing Countries, global production of natural rubber was up in 2017 to ~13.3 million metric tons, but global consumption of NR dropped to ~12.9 million tons [2]. Therefore, research and development to expand commercial utilization of NR is required. NR latex (NRL) is a concentrated colloidal suspension produced by rubber trees. NRL is mainly composed of polyisoprene (poly(2-methyl-1,3-butadiene)). NR is a natural polymer of isoprene and is biodegradable. The most important property of NR is elasticity, which is the ability to return to its original shape and size. However, there are properties of NR that need to be improved, for example, hardness, Young's modulus, and abrasion resistance [3]. Various natural fibers have been used as a reinforcement material in NR matrices such as sisal/oil palm hybrid fibers [4], pineapple fibers [5], coconut fibers [6], bamboo fibers [7], and grass fibers [8].

Bacterial cellulose (BC) is nanocellulose produced by bacteria, principally of genera Acetobacter, such as *Acetobacter xylinum*, in the form of interconnected networks of cellulose nanofibers. BC possesses

unique properties such as high purity (excluding hemicellulose and lignin), high crystallinity, excellent mechanical strength (very high modulus and tensile strength), excellent biodegradability, high water uptake capacity (up to 100 cc/g), and excellent biological affinity [9]. The unique morphological alignment with Nano nonwoven structure of BC resulted in large surface area as compared with plant cellulose fibers or electrospun cellulose nanofibers [10]. With numerous advantageous characteristics, BC has found use across multiple industries. BC is adopted by the paper industry as an emulsion-stabilizing compound. Furthermore, BC can be applied in the medical area as artificial skin for patients with burns and ulcers [11], as artificial blood vessels [12], for drug delivery, and for tissue engineering and wound healing [13].

Although BC possesses numerous useful properties, one disadvantage is the low-breaking elongation. Conversely, NR is well-known for possessing excellent elastic properties. Reinforcements of NR achieved using graphene, carbon nanotubes, or nanocellulose fibers, such as bacterial cellulose, have been previously reported [14–17]. From our previous study, reinforcement of NR with BC was performed via a latex aqueous micro dispersion process and the films of BC–NR by incorporation of BC fibers into NR matrices demonstrated good water affinity and increased mechanical properties when compared with pure NR matrices [17]. However, the composite film of NR incorporated into BC matrices has not been reported so far. Herein, to obtain films with excellent mechanical properties of BC, characterized by high tensile strength, and NR, characterized by high elasticity, NR–BC films possessing a high degree of mechanical strength were developed by immersing BC pellicles into a diluted NRL suspension. Operational parameters such as temperature and NR concentration were studied to optimize the immersion process. NR–BC films were characterized for their chemical and mechanical properties. To the best of our knowledge, this is the first report of such NR–BC films.

#### **2. Materials and Methods**

The *Acetobacter xylinum* (AGR60) was isolated from nata de coco. The stock culture was kindly supplied by Pramote Tammarat, the Institute of Food Research and Product Development, Kasetsart University, Bangkok, Thailand. Natural rubber latex (NRL) with 60% dry rubber content (DRC) was purchased from the Rubber Research Institute of Thailand (RRIT, Bangkok, Thailand). Sucrose and ammonium sulfate were purchased from Ajax Finechem Pty Ltd (New South Wales, Australia). Acetic acid was purchased from Mallinckrodt Chemicals (Paris, KY, USA). Absolute ethanol was purchased from QRec (Chonburi, Thailand).

#### *2.1. Film Preparation*

For BC biosynthesis, the medium for the inoculum was coconut-water supplemented with 5.0% (w/v) sucrose, 0.5% (w/v) ammonium sulfate, and 1.0% (w/v) acetic acid. The medium was sterilized at 110 ◦C for 5 min. Precultures were prepared by a transfer of 50 mL stock culture to 1000 mL in 1500 mL bottle and incubated statically at 30 ◦C for 7 days. After the surface pellicle was removed, a 5% (v/v) preculture broth was added to sterile medium and statically incubated at 30 ◦C for five days in a Petri-dish. All sample BC pellicles were purified by washing with deionized water (DI) for 30 min and, then treated with 1% NaOH (w/v) at room temperature to remove bacterial cells for 24 h, followed by a rinse with water until the pH became 7.0. The BC pellicles were soaked in DI water and stored at 4 ◦C until use.

The procedure for the preparation of BC reinforced with NR (NR–BC) films was developed as follows. NRL was diluted with DI water to form NRL suspension with concentrations of 0.5%–10% dry rubber content (DRC) (expressed as weight per volume). In order to reduce the viscosity of NRL suspension, 6 mL of 50% (v/v) aqueous ethanol solution was slowly added into 300 mL NRL suspension. BC pellicle was then immersed in NRL suspension with concentrations of 0.5%–10% DRC for 48 h as the immersion temperature varied from 30–70 ◦C. Then, it was washed with DI water, air-dried at 30 ◦C for 48 h, and stored in plastic film at room temperature. BC was defined as the unmodified BC and NR–BC was defined as the modified BC by immersing in an NRL suspension. The xNR–BCy

film was defined as the NR–BC film modified by immersing in an NRL suspension at x% DRC and at an immersion temperature of y ◦C. For example, 0.5NR–BC50 was the NR–BC film modified by immersing in an NRL suspension at 0.5% DRC and at a temperature of 50 ◦C.

#### *2.2. Characterization*

#### 2.2.1. Field Emission Scanning Electron Microscopy (FESEM)

The examination of the surface morphology was performed by field emission scanning electron microscopy (FESEM). Scanning electron micrographs were taken with JEOL JSM-7610F microscope (Tokyo, Japan). The films were frozen in liquid nitrogen, immediately snapped, and vacuum-dried. Then, the films were sputtered with gold and photographed. The coated specimens were kept in dry place before the analysis. The FE-SEM was obtained at 10 kV, which was considered to be a suitable condition for these samples. The average thickness of the dried BC films and NR–BC films was measured using the ImageJ program.

#### 2.2.2. Laser Particle Size Distribution (PSD)

The particle sizes of rubber in NRL and NRL added with ethanol were investigated by laser diffraction technique. Particle size distribution curves were taken by Mastersizer 3000 (Malvern, UK). The operating size classes were recorded in the range of 0.01–3000 μm at a stirrer speed of 2000 rpm.

#### 2.2.3. Fourier Transform Infrared Spectroscopy (FTIR)

The chemical structures of the films were analyzed and recorded by FTIR with a Nicolet SX-170 FTIR spectrometer (Thermo Fisher Scientific, Waltham, MA, USA) in the region of 4000–500 cm−<sup>1</sup> at a resolution of 4 cm<sup>−</sup>1.

#### 2.2.4. Water Absorption Capacity (WAC)

Water absorption capacity (WAC) was determined by immersing the pre-weight of dry films of 20 mm × 20 mm in DI water at room temperature (30 ◦C) until equilibrium. Then, the films were removed from water and blotted out with Kim wipes. The weights of the hydrate films were then measured, and the procedure was repeated until there was no further weight change. WAC was calculated by the following Equation (1):

$$\text{WACC}^{\circ}\_{\text{V}} = \left[ \frac{\mathcal{W}\_{h} - \mathcal{W}\_{d}}{\mathcal{W}\_{d}} \right] \times 100,\tag{1}$$

where *Wh* and *Wd* denote the weights of hydrate and dry films, respectively.

#### 2.2.5. Toluene Uptake (TU)

Specimens of BC and NR–BC films of 20 mm × 20 mm were weighed and immersed in toluene at room temperature. After that, the specimens were weighed. The procedure was repeated until there was no further weight change. The toluene uptake (TU) was calculated by the following Equation (2):

$$\text{TU}\% = \left[\frac{W\_t - W\_0}{W\_0}\right] \times 100,\tag{2}$$

where *W0* and *Wt* denote the weights of films before and after the immersion in toluene, respectively.

#### 2.2.6. X-Ray Diffraction (XRD)

The examination of the crystal structures of the films was performed by X-ray diffractometer (model D8 Discover, Bruker AXS, Karlsruhe, Germany). The films were cut into strip-shaped specimens of 4 cm in width and 5 cm in length. The operating voltage and current were 40 kV and 40 mA, respectively. Samples were scanned from 5–40◦ 2θ using CuKa radiation.

#### 2.2.7. Differential Scanning Calorimetry (DSC)

DSC analysis was used to measure the thermal properties of the films, such as glass transition temperature (Tg) and crystalline melting temperature (Tm). A sample of about 4 mg was sealed in aluminum pan for DSC analysis under nitrogen gas. In addition, the curing behavior of the films was determined using a NETZSCH DSC 204 F1 Phoenix (Selb, Germany). The scanning range was −100 to 200 ◦C with a heating rate of 5 ◦C/min.

#### 2.2.8. Thermal Gravimetric Analysis (TGA)

The thermal weight changes of BC, NR, and the films were determined using TGA (Q50 V6.7 Build 203, Universal V4.5A TA Instruments, New Castle, DE, USA) in a nitrogen atmosphere. The scanning range was 30 ◦C to 700 ◦C with a heating rate of 10 ◦C/min. The initial weight of each sample was around 10 mg and percentage weight loss versus decomposition temperature by TGA analysis was determined.

#### 2.2.9. Mechanical Properties Testing

The tensile strength of the films was measured by Instron Testing Machine (ASD8-82A.TSX, NY, USA). The test conditions followed ASTM D882. The determination of elongation at break, tensile strength, and Young's modulus was performed using films in strip-shaped specimens of 10 mm in width and 10 cm in length. The mechanical properties of each sample were the average values determined from five specimens.

#### 2.2.10. Biodegradation in Soil

Biodegradation of BC and NR–BC films in soil for six weeks was evaluated. The samples were cut into square pieces of 3 cm × 3 cm and were buried in 10 cm soil depth under uncontrolled temperature (24–35 ◦C). Samples were taken out every week and washed with DI water. Then, the samples were dried at 50 ◦C for 24 h and weighed. The specific biodegradation rates based on the mass loss of films were determined by the following Equation (3):

$$\text{Biodegradation\%} = \left[\frac{W\_1 - W\_2}{W\_1}\right] \times 100,\tag{3}$$

where *W1* and *W2* denote the initial dry weight of the samples (g) and the residue dry weight of films after biodegradation in soil, respectively.

#### **3. Results**

#### *3.1. NR Particle Size Distribution*

The particle size distribution (PSD) of NR was analyzed by a laser diffraction particle size technique. The rubber particle size from NRL suspension varied from 0.01–2 μm, as shown in Figure 1. A bimodal PSD having two peaks at 0.08 μm and 0.7 μm was observed, with corresponding volume densities of about 2.6% and 9.6%, respectively. A solution of 50% (v/v) ethanol was slowly added into NRL suspension at 2.0% (v/v) to promote the penetration of NR into the BC nanofibrous network structure. No significant change in particle size distribution of NR in NRL suspension with the addition of the ethanol solution was observed.

**Figure 1.** Particle size distribution of natural rubber (NR) in NR latex (NRL) suspension: in the absence of ethanol aqueous solution (solid line) and with the addition of ethanol solution (dash line).

#### *3.2. E*ff*ects of NRL Concentration and Temperature on Integration of NR into BC*

Surface morphology of a never-dried BC film is shown in Figure 2a. Surface and cross section of dried BC films are shown in Figure 2b,c, respectively. According to the observation from the FESEM images, the surface of never-dried BC films comprises microporous structures of fibrous networks with nanocellulose fiber diameters of ~50–100 nm. Never-dried BC films possess pore sizes ranging between 0.1–2 μm located between the fibers. A similar pore structure of unmodified BC hydrogel was previously reported [18]. After air drying at 30 ◦C, a dense nanocellulose layer structure was obtained, as shown in Figure 2b. Water loss during drying could result in shrinkage and a compact structure. The cross section of dried BC films in Figure 2c shows that the structures are composed of multilayers of thin sheets, which is the characteristic feature of BC film previously reported [13]. During the immersion, NR diffused into and gradually filled the never-dried BC pores, and by doing so, might coat the surface. FESEM images of the cross section of the dried NR–BC films by immersion in NRL suspensions of 0.5%, 2.5%, 5%, and 10% DRC at 50 ◦C are shown in Figure 3. The NR–BC films comprise dense nanocellulose layers incorporated with NR. The amount of integration of NR into BC was in association with the increase in weight (Figure 4) and thickness of the dried films (Table 1). The accumulated NR in the composite was estimated from the change of dried weight of the films before and after the imersion in NRL suspension, as shown in Figure 4. The average weight of the dried BC film was 0.0091 ± 0.0001 g, whereas the diameter and thickness were 14.01 ± 0.13 cm and 12.05 ± 0.26 μm, respectively. The diffusion and integration of NR into BC increased as a function of NRL concentration up to around 2.5%–5% DRC. The high amount of NR accumulated in the composite films was obtained by the immersion in NRL suspension at 2.5% DRC (at 50 ◦C), 5% DRC (at 50 ◦C), 2.5% DRC (at 60 ◦C), and 5% DRC (at 60 ◦C), where the estimated ratios of NR/BC in the composite films were 53.9%, 42.9%, 73.6%, and 70.3%, respectively (the amounts of NR in the composite films were 35.0%, 30.0%, 42.4%, and 41.3%, respectively). However, further increasing the NR concentration in the suspension to 7.5% and 10% DRC resulted in a significant decrease of NR adsorption. Higher weights and thicknesses of films were also observed as a function of increased immersion temperature from 30 ◦C to 50–60 ◦C. The rate of diffusion of NR into nanocellulose fiber networks should generally increase with temperature up to a certain point as a result of increased kinetic energy. Similar changes to film thickness and surface area have previously been reported in relation to the modification of BC by impregnation with aqueous alginate solutions at 1%–3% (w/v) at the immersion temperature of 30 ◦C to 50 ◦C [19].

**Figure 2.** Field emission scanning electron microscopy (FESEM) images of surface morphologies of never dried films of bacterial cellulose (BC) (**a**) and dried BC films: surface (**b**) and cross section (**c**).

**Figure 3.** FESEM images of the cross section of dried films of 0.5NR–BC50 (**a**); 2.5NR–BC50 (**b**); 5NR–BC50 (**c**); 10NR–BC50 (**d**); and a closer look of the surface of dried NR–BC film (**e**).

**Figure 4.** Dry weight of NR–BC films modified by immersing in an NRL suspension at various concentrations (0%–10% dry rubber content (DRC)) at immersion temperatures of 30 ◦C, 50 ◦C, 60 ◦C, and 70 ◦C. Values were expressed as mean ±SD (*n* = 3); blue bar = the dry weights of BC before the immersion and white bar = the estimated NR dry weights in the composites.


**Table 1.** Thickness of natural rubber (NR)–bacterial cellulose (BC) films.

Values are expressed as mean ±SD (*n* = 3).

#### *3.3. Mechanical Properties*

The mechanical properties of BC and NR–BC films were analyzed in terms of elongation at break, tensile strength, and Young's modulus, as shown in Figure 5. According to our previous studies, the tensile strength, elongation at break, and Young's modulus of unmodified BC films could be varied around 70–300 MPa, 0.5%–5%, and 5–17 GPa, respectively, depending on many factors, such as culture conditions, drying conditions, and bacteria strain [19–21]. In this study, the unmodified BC film possessed an elongation at break, tensile strength, and Young's modulus of 0.6%, 112.4 MPa, and 9.14 GPa, respectively; meanwhile, the corresponding values for the uncured NR film developed from pure NRL were around 100%–111%, 0.8–1.2 MPa, and 1.6–2.4 MPa, respectively [17,22]. The NR–BC films have considerably higher elongation at break compared with the unmodified BC films. The maximum values of elongation at break, at 3.0%–3.5%, were obtained when immersing BC in NRL at a concentration range of 2.5%–5% DRC between 50 and 60 ◦C. The incorporation of NR into the BC matrices resulted in the ability of the fibers to absorb more energy, and thus prevent the nanocellulose fibers from breaking. Consequently, when compared with the BC films, the films were demonstrated to elongate more prior to breaking. The NR–BC film tensile strengths are also significantly higher when compared with those of the BC-only film (Figure 5b. The optimal conditions to achieve maximum tensile strength of the NR-BC films are during the preparation of 2.5NR–BC50, in which the tensile strength increased to 392.4 MPa (a ~3.5-fold increase over that of the unmodified BC film). Additionally, the films treated with NRL suspension of 5% DRC and at the immersion temperature of 50–60 ◦C exhibited considerably higher tensile strengths when compared with the unmodified BC film. The tensile strengths of 2.5NR–BC50, 2.5NR–BC60, 5NR–BC50, and 5NR–BC60 showed a 3.5-, 1.7-, 2.3-, and 2.0-fold increase, respectively, over that of the unmodified BC film. The Young's modulus of pure BC and the films are shown in Figure 5c. The unmodified BC film exhibits a Young's modulus of 9.14 GPa, whereas the NR–BC films display higher Young's modulus values. Under the optimal preparation conditions, 2.5NR–BC50 displayed a Young's modulus value of 20.05 GPa or an ~2-fold increase over that of the BC film. As the maximum enhancement of the mechanical properties was obtained by immersing at 50 ◦C, the results of the following studies were performed on NR–BC films prepared by immersing in NRL suspensions at 50 ◦C.

**Figure 5.** Elongation at break (**a**), tensile strength (**b**), and Young's modulus (**c**) of the dried BC and NR–BC films modified by the immersion in NRL of various concentrations (0%–10% DRC) at various immersion temperatures.

#### *3.4. FTIR Analysis*

As shown in Figure 6, the FTIR spectrum of pure BC shows characteristic peaks at 3347 cm−<sup>1</sup> attributed to O–H stretching vibrations from hydroxyl groups, and at 2900–2800 cm−<sup>1</sup> corresponding to C–H stretching, while a peak at 1650–1640 cm−<sup>1</sup> corresponds to the vibration of the carbonyl group (C=O). A peak at 1440 cm−<sup>1</sup> is attributed to CH2 bending, and a peak located at 1165–1060 cm−<sup>1</sup> is attributed to C–O stretching [17,23,24]. The FTIR spectrum of NR reveals several characteristic peaks. The peak at 2960 cm−<sup>1</sup> is attributed to the vibration of C–H stretching, while the peak located at 2917 cm−<sup>1</sup> corresponds to the symmetric stretching of methylene (−CH2). The asymmetric stretching of the CH3 group is observed at 2847 cm<sup>−</sup>1. A peak located at 1637 cm−<sup>1</sup> is assigned to the vibration of C=C, while the peak located at 1465 cm−<sup>1</sup> is attributed to the symmetric bending of CH2. The peak at 1375–1450 cm−<sup>1</sup> is attributed to the vibration of CH2 asymmetric bending and stretching [17,25]. The FT-IR spectra of the NR–BC films display characteristic peaks of both BC and NR. No occurrence of new peaks was observed.

**Figure 6.** Fourier transform infrared spectroscopy (FTIR) spectra of BC, NR, and NR–BC films treated by the immersion in NRL of various concentrations at 50 ◦C.

#### *3.5. X-Ray Di*ff*raction (XRD) Analysis*

The XRD patterns of BC and the NR–BC films are shown in Figure 7. The characteristic peaks of BC comprise three main peaks at two theta angles of 14.5◦, 17.0◦, and 22.8◦, associated with the typical profile of cellulose [26]. The diffraction patterns of the NR–BC films exhibit a higher degree of order, as seen by the sharp peaks, whereas the diffraction pattern relating to the BC-only film displays relatively broad peaks, indicating a less crystalline material. The NR film displays a typical diffraction pattern of an amorphous polymer having a prominent broad hump located at a two-theta angle of 18◦. The NR–BC films, prepared at various concentrations and immersion temperatures, display the same characteristics of the BC-only film diffraction pattern, indicating the presence of the BC crystalline structure. The peaks at two-theta ≈ 17◦ and 22.8◦ became slightly more intense with increasing NR, which might imply that the integration of NR into fibrous structure of BC might have some effect on the crystalline structure of BC. The amorphous broad hump associated with NR was not observed in the XRD patterns, as there was a small proportion of NR in the BC matrices.

**Figure 7.** X-ray diffraction (XRD) patterns of BC, NR, and NR–BC films treated by the immersion in NRL of various concentrations at 50 ◦C.

#### *3.6. Di*ff*erential Scanning Calorimetry (DSC)*

The DSC thermograms of dried BC, NR, and the NR–BC films are shown in Figure 8. The glass transition temperature (Tg) of pure BC was barely detected because the highly crystalline structure of nanocellulose exhibited flat heat flow curves, which are difficult to separate from the baseline. For the thermal degration, the BC film displays two exothermic broad peaks. The first peak around 217 ◦C is associated with the thermal degradation of proteinaceous matter in the BC film [19,26]. The second peak at ~304 and 340 ◦C is ascribed to the partial pyrolysis of cellulose [27]. The DSC thermogram of NR exhibited a Tg at −68.1 ◦C. The DSC thermograms of 5.0NR–BC50 and 10.0NR–BC50 exhibited a Tg at around −67 ◦C, slightly higher than the Tg of NR. The endothermic peaks exhibited by the NR–BC films between 30–150 ◦C are attributed to water loss or dehydration of the films [19,28]. NR loading into the BC matrices (2.5NR–BC50, 5.0NR–BC50, and 10.0NR–BC50) resulted in a slight change to the position of the exothermic peaks.

**Figure 8.** Differential scanning calorimetry (DSC) chromatograms of BC (**a**), NR (**f**), and NR–BC films of 0.5 NR–BC50 (**b**), 2.5 NR–BC50 (**c**), 5.0 NR–BC50 (**d**), and 10.0 NR–BC50 (**e**).

#### *3.7. Thermal Gravimetric Analysis (TGA) and Di*ff*erential Thermal Analysis (DTA)*

The TGA and DTA curves of BC, NR, and the NR–BC films are shown in Figure 9a,b, respectively. The slight weight loss from 30–200 ◦C is associated with the vaporization of water. The percentage weight loss of pure BC across a temperature range of 210–240 ◦C is ~6 wt.%, indicating the decomposition of proteins [29]. The major pyrolysis of BC, resulting from the decomposition of cellulose [30], occurs at ~300–360 ◦C, and results in a residual char product (≈20 wt.%). Conversely, the pyrolysis temperature of NR occurs at ~340–440 ◦C, at which about half of the mass loss is observed for pyrolysis at 380 ◦C. The thermal decomposition of the NR–BC films is divided into three weight loss stages. At temperatures <200 ◦C, the weight loss is associated with water loss. The second and third weight loss stages occur across a temperature range of 300–400 ◦C. The maximum rate of weight loss of the films occur at ~320 ◦C and ~380 ◦C in accordance with the decomposition of BC and NR, respectively. Films of 5.0NR–BC50 and 10.0NR–BC50 presented slightly increased thermal stability when compared with the BC film and other NR–BC films.

**Figure 9.** Thermal gravimetric analysis (TGA) (**a**) and differential thermal analysis (DTA) (**b**) curves of BC, NR, and NR–BC films.

#### *3.8. Water Absorption Capacity (WAC)*

Water absorption capacities of BC, NR, and NR–BC films treated by the immersion in NRL of various concentrations at 50 ◦C are shown in Table 2 and Figure 10. Overall, all films showed rapid adsorption of water in the first 20 min and then slow adsorption until concentrations reached equilibrium at around 1 h. The BC film has highly water absorption capacity at 610.5% because of the hydrophilic property of the hydroxyl group in its structure [31,32]. Because of the hydrophobic structure of NR, the water absorption capacity of the NR film was very low (10.9%). As compared with BC, NR–BC films had significantly higher resistance to water.

**Table 2.** Water absorption capacity (WAC%) and toluene uptake (TU%) of NR–BC50 films.


**Figure 10.** Water absorption capacity (WAC %) with time of NR–BC50 films.

#### *3.9. Toluene Uptake (TU)*

NR is a nonpolar material; therefore, it is soluble in non-polar solvents. Toluene is an aromatic solvent that is mostly used in many rubber industries. Thus, the effect of toluene uptake on NR-BC films was investigated (Table 2). At initial absorption, the toluene uptake of the NR film rapidly increased and reached the maximum value at 2642% in 1 h. After that, the degradation of the NR film by dissolving in toluene was observed [33]. Because of its hydrophilic nature, the BC film had high resistance to non-polar solvents and exhibited very low toluene uptake at around 6.4%. The toluene uptake of NR–BC films is higher than BC film, but so much less than NR (0.02–0.05 of NR film).

#### *3.10. Biodegradation in Soil*

Biodegradability is an essential property when considering environment issues, and is a critical property for the application of green packaging materials. BC structures comprise crystalline nanocellulose fibers and a minor amount of amorphous cellulose chains, which can be attacked by multiple microorganisms in the soil through enzymatic degradation [34,35]. The conformation of NR is relatively resistant to biodegradation through microorganisms when compared with many other natural polymers [36]. However, there are known microorganisms in soil such as bacteria and fungi that have the ability to degrade NR [36,37]. Natural latex rubber is biodegradable, as is claimed by numerous products and manufacturers. Natural rubber latex gloves can be disposed of by either landfill or incineration, which are not harmful to the environment. Recently, it was shown that the mixed culture isolated from soil samples collected from rubber contaminated ground in Songkhla province, Thailand had potential in degrading rubber, in which significant changes could be detected within 30 days [38]. In this study, the biodegradability of BC and the NR–BC films in soil is shown in Figures 11 and 12. The films underwent soil burial experiments and the average soil temperature was 35.1 ± 2.0 ◦C. The BC film exhibited a higher weight loss percentage when compared with the other films, and was completely decomposed within four weeks. The NR–BC films were completely decomposed within 4–6 weeks. Overall, higher NR loadings were observed to slow the decomposition rate of the films. Films of 2.5NR–BC50 and 5NR–BC50 demonstrated a higher resistance to microorganism degradation when compared with 0.5NR–BC50 and 10NR–BC50. The films of 2.5NR–BC50 and 5NR–BC50 were completely decomposed within six and five weeks, respectively, whereas 0.5NR–BC50 and 10NR–BC50 were completely decomposed within four weeks.

**Figure 11.** Biodegradability of films in soil after 0–6 weeks. (BC = ×; 0.5 NR–BC50 = white triangle; 2.5 NR–BC50 = white circle; 5 NR–BC50 = black circle; 10 NR–BC50 = black triangle).

**Figure 12.** The images of degraded materials at the different weeks in the study of biodegradation in soil.

#### **4. Discussion**

BC pellicle was immersed in diluted NR latex (NRL) suspensions with the supplement of ethanol aqueous solution. The slow addition of a 50% (v/v) ethanol solution into the NRL suspension at 2.0% (v/v) did not show a significant effect on NR particle size. However, it could result in a decrease in viscosity of the NRL suspension as a result of the reduction of the gel content of rubbers from NRL [39]. It was found that the addition of ethanol in NRL at a specific fraction could promote the penetration of NR into the BC nanofibrous network structure. However, as ethanol is an organic polar solvent, the addition at too high a fraction into the NRL suspension resulted in the coagulation of NR molecules [40]. From our preliminary test, the addition of ethanol at concentration ≥70% v/v into NRL for 2.0% (v/v) or the addition of 50% v/v ethanol into NRL for ≥3.0% v/v caused coagulation of NR.

After the modification by immersing BC films in diluted NRL suspensions, it was found that the maximum dry weight of the NR–BC films was around 0.014–0.016 g, which was about 1.5–1.8 of that of the BC film. The maximum thickness of the NR–BC films was 20–27 μm, or about 1.7–2.3 of that of the BC film. The maximum amount of integration of NR into BC was obtained by immersing BC in NRL suspensions of 2.5%–5.0% DRC at an immersion temperature of 50–60 ◦C. However, further increasing the NRL concentration above 5.0% DRC led to the agglomeration of NR molecules, resulting in a lower diffusion of NR molecules into the BC film. At higher concentrations, the close vicinity of the NR molecules with respect to each other could result in increased interaction and agglomeration. As the BC pellicle pore size was relatively small, the diffusion of agglomerated NR into the BC fibrous network was hindered. Considering how diffusion is influenced by immersion temperature, at low NRL concentrations (< 2.5% DRC), no significant increase of NR–BC film thickness was observed when increasing the immersion temperature from 30 ◦C to 70 ◦C. At low NR concentrations, low adhesion of NR to the BC matrices was observed. When BC was immersed in NRL suspensions at

medium to high concentrations (2.5%–10% DRC), the NR–BC film thickness significantly increased as a function of increased immersion temperature from 30 ◦C to 60 ◦C. Increased kinetic energy of the NR molecules at high temperature was considered to promote the diffusion of NR into the BC matrices. At high temperature, the agitated particles were subjected to stronger and more frequent collisions. However, NR–BC film thickness and weight decreased as a function of elevated immersion temperature from 60 ◦C to 70 ◦C, which implied less accumulation of NR in BC films. When subjected to a high immersion temperature of 70 ◦C, the NR molecules, as a result of a high collision rate, could form particle agglomeration. The coating of agglomerated NR on the BC surface and in the BC fibrous networks was thought to prevent the diffusion of small NR particles into the BC pores. The low degree of NR penetration resulted in a smaller NR–BC film thickness when the immersion temperature increased from 60 ◦C to 70 ◦C.

BC films usually possess a high mechanical strength (high modulus and tensile strength), but low elongation at break (or low fracture strain). Conversely, NR films show higher elongation at break, but possess a relatively lower mechanical strength as compared with BC films. In this study, it was shown that the mechanical properties of the BC films were considerably improved by the addition of NR into BC matrices. The important factors that affect the diffusion of NR into BC matrices are the concentration of NRL suspension and temperature. The concentration of NRL suspension at 2.5%–5.0% DRC and the immersion temperature at 50–60 ◦C are the optimal conditions for high diffusion of NR into BC matrices. The integration of NR into BC matrices resulted in improved mechanical properties of the films. When subjected to a high immersion temperature of 70 ◦C, the NR molecules, as a result of a high collision rate, could form particle agglomeration. The agglomerated NR particle could cover some parts of the BC surface and filled in the pores of BC matrices. The large particles could prevent the diffusion of small NR particles into the pores of the inner part. At high temperature, NR agglomeration could be generated to a greater extent, especially in the condition with higher concentration of NRL suspension. This agglomeration might cause a problem of poor distribution of NR inside the BC matrices. As a result, at an immersion temperature of 70 ◦C, NR–BC films exhibited the largest Young's modulus in the 0.5% DRC group, but exhibited the smallest Young's modulus in the 10% DRC group. However, under the optimal condition, NR integrated into the BC films and was well distributed in BC matrices. NR could bind the nanocellulose fibers together and, consequently, the mechanical properties of the NR–BC films, with respect to their modulus and strength, were enhanced compared with the BC film. NR possessed high structural regularity and typically crystallizes spontaneously when stretched [41]. The NR bonds in the nanocellulose fibrous network restricted the movement of the nanocellulose fibers and enhanced mechanical strength [42]. It was revealed that the presence of NR on the BC matrices in the NR–BC films induced superior mechanical properties such as high elongation at break and high tensile strength. Therefore, the combination of a nanocellulose fibrous network and NR could result in a synergistic effect on the mechanical properties.

The FTIR and XRD results showed that there was no chemical interaction between NR and BC; however, the integration of NR into fibrous structure of BC might improve crystalline structure of BC. At high NR diffusion into BC fibers, NR–BC films exhibit relatively high structural and thermal stability. It was shown that 5.0NR–BC50 and 10.0NR–BC50 presented relatively increased thermal stability when compared with that of the BC film and the other NR–BC films. On the other hand, it was shown that XRD peaks of 5.0NR–BC50 and 10.0NR–BC50 are also slightly sharper than the others. Therefore, the integration of NR into BC matrices at a certain content (an optimal concentration range) might have some positive effects on the crystalline structure of BC. The crystalline structure could affect thermal properties of the composites. The result of TGA residual mass (Figure 9) showed that the remaining mass by the higher order was BC > 0.5 NR–BC50 > 2.5 NR–BC 50 > 10NR–BC50 > 5.0 NR–BC 50 > NR. According to our previous study [17], the char yield of BC was higher than that of NR, and the char yields increased along with the ratio of BC in NR composites. In this study, the ratio of BC/NR by the higher order was 0.5NR–BC50 > 10 NR–BC50 > 5.0NR–BC50 > 2.5NR–BC50 (Figure 4). Compared with the other composite films, the remaining mass of the composite film of 0.5 NR–BC50

was the highest because the ratio of BC/NR of this composite film was higher than the others. For the same reason, the residual of 10.0NR–BC50 was greater than that of 5.0NR–BC50. However, it is noticed that the remaining mass of 2.5 NR–BC50, which has the highest ratio of NR (or the lowest ratio of BC) is quite high when compared with the others. As a result, it was suggested that not only the composition, but also the structure of the composites could also have an effect on the thermal properties.

BC has a hydrophilic structure and NR has a hydrophobic structure; therefore, the values of WAC% decreased with the ratio of NR in the NR–BC composites. The films of 2.5NR–BC50 and 5NR–BC50 had the highest water resistance (the lowest WAC%). Because of NR binding into BC fibers, only small amounts of water could diffuse or be adsorbed into NR–BC films. Consequently, the water resistance of NR–BC films increased with NR concentration in the films. The toluene uptake (TU%) of NR–BC50 films was in range of 30%–70% and the increase in toluene uptake was related to the concentration of NR in the films. However, the toluene uptake of NR–BC films was much lower than that of NR films and, after the immersion of NR–BC films in toluene for 4 h, no significant change in overall outlook of the NR–BC films was observed. The good resistance to toluene should be attributed to the hydrophilic nature and high stability of nanocellulose network structure of BC in nonpolar solvents.

The evaluation of biodegradation capability of NR–BC50 films in comparison with BC films was conducted in soil environment for six weeks. BC fiber is biodegradable by various bacteria and fungi in soil. The microorganism first attacks the nanocellulose amorphous region and, thereafter, decomposes all the crystalline regions. On the other hand, it has been previously reported for a slow process of biodegradation of NR and related compounds by some microorganisms in soil [43]. In this study, during the biodegradation test, the NR–BC films transitioned to a loose structure. The films were lumpy and highly stretched as a result of the decomposition of nanocellulose by the microorganisms, while the NR particles in the NR–BC films underwent a relatively slower rate of decomposition than the nanocellulose of BC. However, all NR–BC films were biodegradable and could be degraded completely in soil environment within 5–6 weeks.

#### **5. Conclusions**

Films of BC reinforced with NR, prepared by immersing BC into a diluted NRL suspension, demonstrated superior mechanical properties when compared with BC-only films. The combination of a nanocellulose fibrous network and NR polymer synergistically improves the film mechanical properties. Films of 2.5NR-BC50 demonstrated considerably enhanced tensile strength and elongation at break. The NR–BC films also exhibit high structural and thermal stability and are completely degraded in soil within 5–6 weeks.

**Author Contributions:** Conceptualization, M.P.; Methodology, K.P. and M.P.; Validation, M.P.; Formal Analysis, K.P. and M.P.; Investigation, K.P. and M.P.; Resources, M.P.; Writing—Original Draft Preparation, K.P. and M.P.; Writing—Review & Editing, M.P.; Supervision, M.P.; Project Administration, M.P.; Funding Acquisition, M.P.

**Funding:** This research was funded by Thailand Science Research and Innovation, grant number RGU6280004. **Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Article*
