NC State
Narkpiban, K., Sakdaronnarong, C., Nimchua, T., Pinmanee, P., Thongkred, P., and Poonsawat, T. (2019). "The effect of mechano-enzymatic treatment on the characteristics of cellulose nanofiber obtained from kenaf (Hibiscus cannabinus L.) bark," BioRes. 14(1), 99-119.


Cellulose nanofiber (CNF) was successfully isolated from kenaf bark by microfluidization at 20,000 psi for 40 passes. The combination of hydrothermal process and xylanase treatment prior to CNF isolation led to effective cellulose purification. The fiber used for enzymatic pretreatment for CNF isolation had an 85.9% whiteness index and 85.1% cellulose content. The crystallinity of the cellulose extracted from the kenaf bark continued to increase with successive treatments, as indicated by X-ray diffraction analysis. In addition, the enzyme-treated fiber showed increased thermal stability, as shown by thermogravimetric analysis. After CNF isolation, morphological characterization of the CNF was performed via field emission-scanning electron microscopy and transmission electron microscopy. The CNF had an average diameter that ranged from 5 to 10 nm and no undesired elemental contamination, as evidenced by energy dispersive X-ray spectroscopy. The mechano-enzymatic treatments used in this work to obtain CNF were judged to be a promising technique for the fabrication of biomedical and other high-value materials.

Download PDF

Full Article

The Effect of Mechano-enzymatic Treatment on the Characteristics of Cellulose Nanofiber Obtained from Kenaf (Hibiscus cannabinus L.) Bark

Koranat Narkpiban,a Chularat Sakdaronnarong,b Thidarat Nimchua,c Phitsanu Pinmanee,c Paweena Thongkred,c and Thitivara Poonsawat d,*

Cellulose nanofiber (CNF) was successfully isolated from kenaf bark by microfluidization at 20,000 psi for 40 passes. The combination of hydrothermal process and xylanase treatment prior to CNF isolation led to effective cellulose purification. The fiber used for enzymatic pretreatment for CNF isolation had an 85.9% whiteness index and 85.1% cellulose content. The crystallinity of the cellulose extracted from the kenaf bark continued to increase with successive treatments, as indicated by X-ray diffraction analysis. In addition, the enzyme-treated fiber showed increased thermal stability, as shown by thermogravimetric analysis. After CNF isolation, morphological characterization of the CNF was performed via field emission-scanning electron microscopy and transmission electron microscopy. The CNF had an average diameter that ranged from 5 to 10 nm and no undesired elemental contamination, as evidenced by energy dispersive X-ray spectroscopy. The mechano-enzymatic treatments used in this work to obtain CNF were judged to be a promising technique for the fabrication of biomedical and other high-value materials.

Keywords: Cellulose nanofiber; Hydrothermal treatment; Kenaf; Mechano-enzymatic; Microfluidization; Xylanase

Contact information: a: Bioproduct Science, Department of Science, Faculty of Liberal Art and Science, Kasetsart University, KamphaengSaen, NakhonPathom 73140, Thailand; b: Department of Chemical Engineering, Faculty of Engineering, Mahidol University, 25/25 Phutthamonthon 4 Road, Salaya, Phutthamonthon, NakornPathom 73170, Thailand; c: Enzyme Technology Laboratory, Microbial Biotechnology and Biochemicals Research Unit, National Center for Genetic Engineering and Biotechnology (BIOTEC), 113 Thailand Science Park, Phahonyothin Road, KhlongNueng, KhlongLuang, PathumThani 12120, Thailand; d: Department of Botany, Faculty of Liberal Arts and Science, Kasetsart University, Kamphaeng Saen, Nakhon Pathom 73140, Thailand;

* Corresponding author:


Currently, research on the development of biodegradable nanomaterials from renewable resources is a challenging task because of the growing demand in the biocomposites (Alemdar and Sain 2008; Bandyopadhyay-Ghosh et al. 2010), biomedical materials (Lin and Dufresne 2014; Liang et al. 2017), and electronic substrates industries (Sabo et al. 2016). Nanocellulose is obtained from the cellulose from plants or bacteria (Bandyopadhyay-Ghosh et al. 2010; Lin and Dufresne 2014) and shows several superior properties, such as high surface area, unique optical properties, nanoscale dimensions, high crystallinity, and high strength and stiffness, together with the renewability and biocompatibility of cellulose (Oun and Rhim 2016). Nanocellulose is commonly prepared by subjecting the plant lignocellulosic materials to treatment and delignification processes using single or combined mechanical, chemical, and enzymatic methods to disrupt cellulose-hemicellulose-lignin complexes (Agbor et al. 2011).

In general, nanocellulose from plants can be separated into two categories: cellulose nanofiber (CNF) and cellulose nanocrystal (CNC). The CNF has a rope-like shape and consists of long, flexible nanostrings 5 to 60 nm in diameter with a length on the micrometer scale, formed by alternating crystalline and amorphous domains (Zhao et al. 2013; Saelee et al. 2016). The CNC consists of needlelike cellulose crystals 10 to 20 nm in width and several hundred nanometers in length and has limited flexibility compared to CNF because of its high crystallinity (Rahimi Kord Sofla et al. 2016). The CNCs are often generated via intense acid hydrolysis, while CNFs can be generated by various mechanical processes (Martelli-Tosi et al. 2016), such as refining or high-shear homogenization (Kaushik and Singh 2011), high-pressure homogenization (Li et al. 2014), microfluidization (Liu et al. 2017), ultrafine grinding (Hassan et al. 2014), cryocrushing (Alemdar and Sain 2008), and ultrasonication (Khawas and Deka 2016; Xie et al. 2016). These mechanical techniques provide sufficient shear forces to separate cellulose fibers along the longitudinal axis and assist in separating microfibrillated cellulose from plant fibers (Hassan et al. 2014). A major barrier that must be overcome for the successful commercialization of CNF is the time and energy consumption, as high energy is required for the mechanical disintegration of the starting cellulose macrofibers into nanofibers (Xie et al. 2016). Treatment processes, such as hydrothermal and enzymatic hydrolysis, need to be performed prior to the mechanical processes to lower the process costs, minimize the operating time, and reduce the environmental impacts of the energy use.

Hydrothermal treatment (HTP) is a physicochemical process and is known as the most effective treatment method in terms of operational simplicity, with mild reaction conditions, low costs, and environmental friendliness (Gan et al. 2015). The HTP can break down lignocellulosic structural components (lignin, cellulose, pectin, and hemicellulose) to open up the fiber structure by using liquid water under elevated temperature and pressure (Patel et al. 2016), thus enabling accessibility to the polysaccharide components of the raw lignocellulose. In addition, this treatment helps to partially solubilize the hemicelluloses and separate the lignin structure, resulting in a larger crystalline region and a higher specific surface area (Xiao et al. 2011; Ma et al. 2013). Generally, the HTP is influenced by several factors, including temperature, time, and solid-to-liquor ratio. Among these factors, temperature is considered the most important parameter in the process due to the chemical compositions of the fiber that are degraded at different temperatures. The temperature during processing can vary in a wide range, between 100 °C and 250 °C (Fan et al. 2016). During HTP, the acetyl ester bonds of xylan in hemicellulose are hydrolyzed to yield acetic acid as a byproduct, which could also serve as a catalyst to encourage hemicellulose degradation and acquisition. Consequently, the oligosaccharides in the solution are further hydrolyzed to sugar monomers as products of degradation (Agbor et al. 2011; Xiao et al. 2011). Lignin depolymerization and solubilization partially occur during the HTP, but the complete delignification of lignocellulosic material is not possible because of the recondensation of soluble components originating from lignin (Alvira et al. 2010). Hence, the remaining fraction of lignin in the cellulose residue obtained from the HTP process must be removed by the process of chemical bleaching, which conventionally employs a large amount of chlorine and chlorine-based chemicals. However, the use of such chemicals liberates several toxic materials and polluting products that contaminate the bleaching effluent, including absorbable organic halides (AOX) (Gangwar et al. 2016; Pei et al. 2016).

A recent alternative technique with greater prospects for long-term consequences is xylanase treatment; this technique has been used before the bleaching process to decrease the use of chemicals (Gangwar et al. 2016; Saelee et al. 2016). The beneficial effect of xylanase during the delignification process is the selective hydrolysis of xylan reprecipitated on the fiber surface, which improves the permeability of the fiber to the bleaching chemicals (Roncero et al. 2005; Gangwar et al. 2016). Moreover, xylanase treatment increases the extractability of the chromophores associated with lignin-carbohydrate complexes (LCCs) and the hydrophobic materials from the fibers, as indicated by an increase in fiber brightness (Shatalov and Pereira 2009; Gangwar et al. 2016). The application of xylanases as bleach boosting agents has been known to save chemical costs without interfering with the existing process and also allows for a remarkable reduction in the bleaching chemicals and minimizes environmental impact (Gangwar et al. 2016)

Currently, the utilization of annual plants and agricultural residues, including cotton, banana, flax, hemp, sisal, jute, and kenaf for the production of CNF has been a topic of intense research in various studies (Marrakchi et al. 2011). Kenaf (Hibiscus cannabinus L.) is one of the annual plants that is considered a feasible source of CNF with economic and ecological benefits.

Kenaf consists of two types of fibers: the core and the bast fibers. The core fiber is the inner layer of a fiber (70% of dry weight), while the bast fiber refers to the outer layer of the plant fiber (30% of dry weight). Bast fiber has been reported to have excellent mechanical and physical characteristics for use in generating CNF (Jonoobi et al. 2009; Karimi et al. 2014; Sulaiman et al. 2015). However, although there are many treatment methods for the separation of CNF from kenaf, few are based on economic and environmental considerations. Thus, many studies have focused on using a combination of enzymatic and mechanical treatment methods to reduce both energy and chemical use during CNF isolation (Siddiqui et al. 2010; Janardhnan and Sain 2011; Kalia et al. 2013; Hassan et al. 2014; Saelee et al. 2016).

For ‘green’ CNF production, microfluidization is one of the mechanical methods widely used to produce CNF and involves the cellulosic materials through an intensifier pump that generates forces of up to 40,000 psi (276 MPa), followed by an interaction chamber that defibrillates the fibers by shear forces at a constant shear rate (Saelee et al. 2016; Liu et al. 2017). This repeatable process results in tiny particles with a uniform distribution. However, using only microfluidization has a negative effect, as it decreases the crystallinity and damages the fiber structure, and energy consumption increases greatly with increasing treatment time (Henriksson et al. 2007). In addition, CNF produced from microfluidization alone tends to aggregate and cause extensive clogging, which may be the major limitation in the application of CNF (Abdul Khalil et al. 2016). To address this issue, enzymatic treatment of lignocellulosic materials needs to be optimally combined with microfluidization to decrease flocculation and clogging and consequently lower the energy consumption of the CNF isolation process. Consequently, this work aims to find an ecofriendly alternative method for the isolation of nanometer-sized single fibers of cellulose, or CNF, from kenaf bark. To the best of the authors’ knowledge, this work is the first to use kenaf bark to isolate CNF via enzymatic treatment and mechanical processes to reduce the amount of chemical consumption involved.



Kenaf (Hibiscus cannabinus, var Khon-Kaen 60) bark was obtained from the Khon-Kaen Agronomy Research Center, Khon-Kaen, Thailand. Kenaf bark was milled into particle sizes between 125 to 180 μm. The xylanase from Aspergillus sp. was used in the enzymatic treatment and was prepared at the Enzyme Technology Laboratory, National Center for Genetic Engineering and Biotechnology (BIOTEC) in Pathum Thani, Thailand. The chemical agents used in the bleaching process, including sodium chlorite (NaClO2) and glacial acetic acid, were supplied by Merck (Darmstadt, Germany) and were of analytical grade.


Xylanase concentration on hydrolysis of the kenaf bark

To determine the effect of xylanase on fiber degradation, bast tissue from the stem of kenaf was manually separated and milled into 125 to 180 m lengths. Approximately 0.15 g dry weight (DW) of bast section was transferred to a 5-mL centrifuge tube containing 1.5 mL of 50 mM sodium acetate buffer at pH 5.5. The mixture was incubated overnight at 30 °C under agitation (220 rpm). After an equilibration period, xylanase was added to reach activities of 20, 50, 500, and 5000 U/g fiber. One unit (U) of enzyme activity was the amount of xylanase activity necessary to release 1 µmole of xylose equivalent per min at 50 °C. The reaction mixture was incubated at 50 °C under agitation (220 rpm). The samples were taken after 24 h of incubation. Subsequently, the mixed liquid was centrifuged at 12,000 rpm for 10 min and filtered through a syringe filter (13 mm NYL 0.2 µm, Xiboshi, Tianjin Fuji Science & Technology Co., Ltd., Tianjin, China) before analysis of the total reducing sugar by the dinitrosalicylic acid (DNS) method (Miller 1959). All experiments were conducted in triplicate, and their results are given as the mean and standard deviation.

Hydrothermal treatment of kenaf bark

The hydrothermal treatment of kenaf bark was performed in a high-pressure reactor (Parr 4848 reactor controller, Parr Instrument Company, Moline, IL, USA) at 160 °C for 40 min with a fixed fiber to liquor ratio of 1:10. Subsequently, the reactor was cooled immediately, and the hydrothermally pretreated kenaf fiber was obtained via filtration, then oven-dried overnight, and stored for later enzymatic treatment.

Enzymatic treatment of hydrothermally pretreated kenaf bark

The hydrothermally pretreated kenaf fibers were mixed with a xylanase concentration of 5000 U/g fiber at pH 5.5 using a fiber-liquid ratio of 1:10, where the control was the kenaf fiber and buffer (NaAc buffer). The reaction mixture was further incubated at 50 °C with constant agitation (220 rpm) for 24 h. After completion of the enzymatic treatment, the sample fibers were washed well with deionized water and dried at 70 °C for 24 h.

Bleaching process

The bleaching process was adapted from Saelee et al. (2016). Briefly, a mixture of hydrothermally pretreated fibers and bleaching solution (NaClO2 at 2% w/v, adjusted to pH 3.0 with acetic acid) was added to 250-mL Erlenmeyer flasks with a fiber-to-liquid ratio of 1:10. The bleaching process took place at 70 °C in a water bath. A fresh bleaching solution was replaced hourly at the same loading ratio, this was done 4 times. After 4 h of bleaching, the bleached fibers were washed with deionized water until the pH became neutral. Finally, the solid residue was dried in a hot air oven at 60 °C for 24 h.

Preparation of CNF from bleached kenaf fiber

The CNF was produced from kenaf fiber after sequential hydrothermal treatment, enzymatic treatment, and bleaching (HT-KF-x5000-B, Table 1) using a microfluidizer (Microfluidizer M-110P; Microfluidics Corp., Westwood, MA, USA). Diluted suspensions (0.2% w/v) of the HT-KF-x5000-B kenaf sample were prepared and soaked in distilled water for 24 h. The suspensions were processed in the microfluidizer with an 87-m sized at 20,000 psi for 40 passes. Finally, the CNF was collected via centrifugation at 5,000 rpm for 15 min and kept in aqueous solution in the refrigerator until use. The list of fiber sample acronyms is given in Table 1.

Table 1. List of Fiber Acronyms at Different Stages of Treatment in the CNF Isolation Process


Chemical composition of kenaf fiber

The chemical compositions of raw kenaf bark and the fibers obtained after each process were analyzed according to the method developed by Van Soest (1963). The amounts of neutral detergent fiber (NDF), acid detergent fiber (ADF), and acid detergent lignin (ADL) were determined via a fiber analyzer (ANKOM A200; ANKOM Technology, Macedon, NY, USA) using a gravimetric analysis technique. Each sample was measured in triplicate. The hemicellulose, cellulose, and lignin contents were calculated as follows according to Eqs. 1 through 3:

Hemicellulose = NDF – ADF (1)

Cellulose = ADF – ADL (2)

Lignin = ADL (3)

Color analysis of kenaf fiber

The changes in fiber color were evaluated by a spectrophotometer (Datacolor 650TM; Datacolor, Lawrenceville, NJ, USA). A white standard plate was utilized to calibrate the device prior to color measurement (L = 96.37, a = -0.1, and b = 1.33). The color values are shown in three coordinate dimensions: L* lightness (0 = black, 100 = white), a* redness/greenness (-a* = green; +a* = red), and b* yellowness/blueness (-b* = blue; +b* = yellow), including the total color difference (E) value. Each sample was measured in triplicate. The whiteness index (WI) value was computed based on Eq. 4 according to Ghanbarzadeh et al. (2010).

WI = 100 – [(100 – L)2 + a2 + b2]0.5 (4)

Field emission-scanning electron microscopy (FE-SEM)

The surface morphology of the kenaf fiber and CNF was investigated using field emission-scanning electron microscopy (Hitachi Schottky FE-SEM: SU5000 with EM wizard, Tokyo, Japan). The specimen was deposited on the electron microscope grids, and FE-SEM was performed at an accelerating voltage of 10 kV for the kenaf fiber and of 5 kV for CNF in low- and high-vacuum modes.

Transmission electron microscopy (TEM)

The CNF morphology was analyzed under a transmission electron microscope (Hitachi model HT7700; Tokyo, Japan) with an accelerating voltage of 80 kV. The CNF suspension was diluted and sonicated for 30 min. A carbon-coated copper grid was attached above the surface of the diluted CNF droplet. Then, the grid was inverted, allowed to contact a drop of 2% (w/v) uranyl acetate solution for 5 min, and then air-dried at room temperature. The CNF diameter was measured using the software Fiji/ImageJ-win64 (National Institutes of Health, version 1.51d, Bethesda, MD, USA). The average diameter and length of the CNF were determined from 150 fibers in the TEM images.

Energy dispersive X-ray spectroscopy (EDS)

The elemental composition of the sample was analyzed using the EDS technique (Model: RTEM, EDAX Inc., Mahwah, NJ, USA) in combination with TEM. The EDS spectrum obtained from the spot profile revealed the elemental distribution in the sample.

Fourier-transform infrared spectroscopy (FTIR)

The spectra of the fibers after processing were obtained using an FTIR spectrometer (Nicolet 6700; Thermo Scientific, Madison, WI, USA). Each sample was blended with KBr powder, and the mixture was compressed to form a disk. The IR spectra of the fiber samples were measured in transmission mode at 4 cm-1 resolution with 64 scans in a scan range of 4000 to 400 cm-1 with a deuterated triglycine sulfate (DTGS) detector.

X-ray diffraction (XRD)

The XRD patterns of the fibers after each process were obtained by using a high-resolution X-ray diffractometer (PANalytical, X’pert PRO, Almelo, Netherlands). Scattered radiation was detected at a 2 angle from 5° to 60° at a scanning rate of 0.02°/0.5 s scanning rate. The crystallinity index (CrI) was defined by following the XRD peak height method, where the CrI was computed from the height of the 002 peak (I002) and the height of the minimum (Iam) between the 002 and 001 peak according to Eq. 5 (Segal et al. 1959),

CrI (%) = [(I002 – Iam) / I002 ] × 100 (5)

where CrI (%) represents the relative degree of crystallinity, I002 is the intensity of both the crystalline and amorphous regions, and Iam is the intensity of the amorphous region alone.

Atomic force microscopy (AFM)

The morphology of the CNF after microfluidization was observed by AFM (Asylum, MFP-3D AFM (bio), Oxford Instruments, Santa Barbara, CA, USA) at room temperature. The dilute CNF suspension was dropped onto the surface of an optical glass. Then, the sample was air-dried at room temperature and subsequently analyzed.

Thermogravimetric analysis (TGA)

The thermal stability and decomposition temperature of all samples after treatment were determined using a TGA/851e/1600 thermogravimetric analyzer (Mettler Toledo, Schwerzenbach, Switzerland). The samples were heated from 27 °C (room temperature) to 500 °C at a heating rate of 10 °C/min in an ambient nitrogen atmosphere.


Effect of Xylanase Concentration on the Hydrolysis of Kenaf Bark

The effect of xylanase concentration on kenaf bark hydrolysis was investigated by monitoring the amount of total reducing sugars released. The hydrolysis of hemicellulose in kenaf bark increased with the concentration of enzyme used for the treatment (Fig. 1).

Fig. 1. The effect of xylanase at various concentrations (20, 50, 500, and 5000 U/g biomass) on kenaf bark hydrolysis at 50 °C for 24 h (mean value ± standard deviation are shown)

The amounts of reducing sugars obtained after enzymatic hydrolysis with 20, 50, 500, and 5,000 U/g of xylanase activity at 50 °C and pH 5.5 for 24 h were 23.74, 33.03, 99.59, and 360.01 mg/g fiber, respectively. The enzyme hydrolysis at 5,000 U/g clearly gave a maximum amount of reducing sugars, 3.61-, 10.90-, and 15.16-fold greater than the xylanase treatment at the concentrations of 500, 50, and 20 U/g, respectively. In this experiment, xylanase was not purified that contained pectinase activity (data not shown). Therefore, the preparation of kenaf fiber by retting process was skipped. Either pectic substance, a main component of middle lamella, or xylose was removed from kenaf bark in this process. Thus, 5,000 U/g of xylanase is considered. As relationship between enzyme concentrations and substrates in terms of catalytic activity, xylanase activity is defined as follows: “One unit (U) of enzyme activity was the amount of xylanase activity necessary to release 1 µmole of xylose equivalent per min at 50 °C.” Thus, in 1 hour, 20 U/g of xylanase can liberate 1,200 micromoles of xylose, while 5000 U/g of xylanase can liberate 300,000 micromoles of xylose. Therefore, excess xylan in the sample was suitable for the high enzyme activity of xylanase. The concentration of 5000 U/g. xylanase was selected for further investigation on kenaf fiber peeling after hydrothermal treatment.

Effect of Treatment on the Chemical Composition of Kenaf Fibers

The chemical compositions of the fiber after processing are presented in Fig. 2. The percentages of cellulose, hemicellulose, and lignin in the KB specimens were 57.6%, 22.7%, and 2.87%, respectively. After hydrothermal treatment, the cellulose and lignin contents increased from 57.6% and 2.87% to 69.2% and 6.29%, respectively, whereas the hemicellulose content decreased from 22.7% to 14.7% because the hemicelluloses were hydrolyzed into water-soluble sugars. In addition, pectin and acid-soluble lignin were partially removed from the sample, which facilitated the reduction of acid-insoluble lignin after hydrothermal processing (Zhang et al. 2016). After subsequent hydrothermal, xylanase, and bleaching treatments of the kenaf bark, the cellulose content of HT-KF-X5000-B increased from 57.6% to 85.1%, whereas the hemicellulose content declined from 22.7% to 4.05%, and the lignin decreased from 2.87% to 0.15%. The increase in cellulose content was always accompanied by a decrease in the hemicellulose and lignin. This change in chemical composition confirmed that noncellulosic moieties were effectively removed during the treatments.

Fig. 2. Chemical compositions (mean value ± standard deviation) of the fiber at different stages of treatment

Effect of Treatment on the Color of Kenaf Fibers

The whiteness index (WI) of the fiber samples after different treatments indicated the color changes after each treatment, and the appearance of the fibers is illustrated in Fig. 3. The native kenaf bark, or KB, (Fig. 3a) was a light brown color, and the WI value was 62.9. After hydrothermal treatment, the HT-KF fiber (Fig. 3b) was a dark brown color, and the WI value decreased to 50.0. The darker color of the KB after hydrothermal treatment was due to changes in the lignin or its removal and/or dislocation during the hydrothermal process, in addition to the recondensation of the solubilized lignin onto the fiber surface. Thus, the amount of residual lignin after hydrothermal treatment increased from the amount of starting lignin in the raw fiber (Kumar et al. 2013; Saelee et al. 2016). After xylanase treatment, the WI values of the HT-KF-X5000 fibers (Fig. 3d) did not change when compared to the WI values of the HT-KF fibers. Therefore, the hemicellulose dissolution did not affect the whiteness of the fiber. For the control of the xylanase treatment, the HT-KF-NaAc sample (Fig. 3c) had similar WI to the HT-KF sample. After bleaching, the HT-KF-X5000-B fiber (Fig. 3f) was completely white, with a WI of 85.9, due to bleaching agents that either oxidized or reduced the lignin-derived chromophores. The xylanase treatment control, HT-KF-NaAc buffer-B fibers, had a slightly lower WI, 82.15, compared to that of the xylanase-treated fiber, as shown in Fig. 3e. The HT-KF-X5000-B fiber was brighter than the HT-KF-NaAc buffer-B fiber. This indicated that the HT-KF-NaAc buffer-B fiber need more bleaching chemicals than the HT-KF-X5000-B fiber. Consequently, the effects of the xylanase treatment on the kenaf fiber were substantially clearer after the bleaching process.

This result indicated that the xylanase treatment prior to the bleaching processes was effective, as it facilitated an increase in the brightness of the fibers. The reason was that xylanase treatment assisted the bleaching chemical in penetrating the fibers easily for lignin removal, and thus, less of the bleaching chemical was required. These findings were in good agreement with previously reported results regarding the efficiency of xylanase-assisted treatment for the isolation of cellulose nanofibrils from sugarcane bagasse (Saelee et al. 2016).

In general, color is an important key of cellulose fiber because it is used by the customer in many applications. According to the results of color changes, the HT160-KF-X5000-B fiber generated satisfactory color. Thus, the HT160-KF-X5000-B fiber was selected for CNF isolation.

Fig. 3. Images of (a) KB, (b) HT-KF, (c) HT-KF-NaAc buffer, (d) HT-KF-x5000, (e) HT-KF-NaAc buffer-B, and (f) HT-KF-x5000-B

Electron Microscopy Analyses (FE-SEM, TEM, and AFM)

Figure 4 (4a to 4f) shows the fiber surface at each stage of CNF isolation. The FE-SEM image of the KB fiber surface (Fig. 4a) shows the materials covering the fiber bundles that link an individual single fiber into the macro structure. Noncellulosic constituents, such as lignin and hemicellulose, as well as other impurities, such as wax and pectin, were also observed (Karimi et al. 2014; Sulaiman et al. 2015). After hydrothermal treatment at 160 °C, the surface topography of the HT-KF (Fig. 4b) was less smooth because the noncellulosic constituents were removed. This was consistent with the chemical composition of the hydrothermally-treated fiber (Fig. 2). The FE-SEM images showed that the HT-KF-x5000 specimens (Fig. 6d) were less aggregated than the HT-KF-NaAc buffer specimens (Fig. 4c). Therefore, the xylanase treatment facilitated the better separation of the fiber from the bundles. After bleaching, the HT-KF-x5000-B specimens (Fig. 4f) were more disrupted than the HT-KF-NaAc buffer-B specimens (Fig. 4e). This result implied that the enzyme treatment and bleaching process substantially affected the microstructure of the kenaf fibers, which may not be detected by low-magnification imaging and chemical composition analysis. However, the efficiency of the enzyme treatment may change the internal microstructure of the crystallinity of cellulose and the matrix without changing the major components. Thus, the enzyme-treated samples allowed for greater penetration than the untreated samples. Figures 4g and 4h show the FE-SEM micrographs of CNF from the HT-KF-x5000-B suspension after microfluidization and after freeze-drying, respectively. A complete disintegration of cellulose fiber was found after microfluidization.

Fig. 4. FE-SEM images of (a) KB, (b) HT-KF, (c) HT-KF-NaAc buffer, (d) HT-KF-x5000, (e) HT-KF-NaAc buffer-B, (f) HT-KF-x5000-B, (g) CNF suspension after microfluidization, and (h) CNF after freeze-drying

As shown in Fig. 5, the microfluidization process led to the formation of nanofibers, as indicated by the nanodimensional scale of the objects observed in the TEM and AFM images. The TEM images of CNFs at different scales are illustrated in Fig. 5a and 5b. These images revealed CNF filaments form a web-like structure. The diameter distribution of CNF is shown in Fig. 5c. More than 80% of CNFs are 5 to 15 nm in diameter, and their length was several micrometers. Moreover, the nanometer dimensions of CNF were also supported by AFM imaging (Fig. 5d). Therefore, these results indicated that the CNF obtained after microfluidization at 20,000 psi for 40 passes was completely isolated on a nanometer scale.

EDS Analysis

The EDS spectrum of CNF shows the energy levels of carbon and oxygen in the sample, as shown in Fig. 6. The amount of carbon was higher than the amount of oxygen: the composition was 71.09% carbon and 28.91% oxygen. The chemicals applied in several treatment steps, especially sulfur, sodium, and chlorine, used in the xylanase treatment and bleaching process, were not found in the final CNF.

Fig. 5. The morphology of CNF: (a and b) TEM images at different magnifications, (c) diameter distribution, and (d) AFM image of CNF

Fig. 6. EDS spectrum of kenaf CNF

The results were in contrast to those of a previous report, as the chemical residues in the final CNF, such as sulfur and sodium, were present in significant amounts (Rahimi Kord Sofla et al. 2016). Based on these results, the CNF-based materials derived from the present work can be concluded as suitable for biomedical applications, because high-purity CNF is required for safe use (Liang et al. 2017). Changing the bleaching agent from NaClO2 to H2O2 would be interesting to investigate in the future as a means of achieving safer and more environmentally friendly CNF production (Martelli-Tosi et al. 2016).

FTIR Spectroscopy Analysis

To analyze the changes in the chemical structure of fiber bundles, fibers, and CNF after various treatments, FTIR spectroscopy was used. Figure 7 shows the FTIR spectra of (a) KB, (b) HT-KF, (c) HT-KF-NaAc buffer, (d) HT-KF-x5000, (e) HT-KF-NaAc buffer-B, (f) HT-KF-x5000-B, and (g) CNF specimens. The absorbance peaks at 3500 to 3200 cm-1 found in the spectra of all samples were attributed to the O-H stretching and bending vibrations of the hydrogen-bonded hydroxyl groups, and the absorption in the range of 2900 to 2800 cm-1 was assigned to the C-H stretching vibration of the alkyl groups in the aliphatic bonds of cellulose (Karimi et al. 2014; Chandra et al. 2016; Saelee et al. 2016). The peaks located at wavenumber 1640 cm-1, related to the O-H bending vibrations of the adsorbed water in a cellulose fiber structure (Karimi et al. 2014), could be found for all samples. The peak at 1600 cm-1 was associated with functional groups such as methoxyl-O-CH3, C-O-C, and aromatic C=C (Abraham et al. 2013; Chandra et al. 2016). This 1600 cm-1 peak was absent in the spectrum of bleached fiber and CNF because the lignin content was completely removed after the bleaching process.

The absorbance at 1506 cm-1 of the KB, HT-KF, HT-KF- NaAc buffer, and HT-KF-x5000 fiber spectra was associated with the C=C stretching vibration of the aromatic ring of lignin (Xu et al. 2015). This characteristic peak at wavenumber 1506 cm-1 disappeared completely in the spectra of specimens of bleached fiber (HT-KF-NaAc buffer-B and HT-KF-x5000-B) and CNF; however, it was present in the spectra of the unbleached samples (HT-KF-NaAc buffer and HT-KF-x5000). Moreover, the peak at wavenumber 1460 cm-1 corresponded to the C-H bending patterns of the lignin and hemicellulose as well as the C=C vibrations of the aromatic structure in lignin. This peak disappeared from the spectrum of bleached fiber as well as from the spectra of the CNF. It was concluded that the bleaching process removed most of the hemicellulose and lignin from the kenaf fiber and CNF.

For polysaccharide absorption, the absorbance between 1430 to 1420 cm-1 and 1380 to 1330 cm-1 in all samples was associated with the symmetric bending of CH2 and C-O bonds in the polysaccharide aromatic ring (Jonoobi et al. 2009). Moreover, at 1032 cm-1, the wavenumber peaks of all samples were associated with the C-O, C-C stretching, or C-OH bending of cellulose and hemicellulose (Xu et al. 2015). This peak was completely absent from the spectra of the CNF specimens, which indicated that the end product was free of hemicellulose. In addition, the increase in absorbance at 898 cm-1 that corresponded to the C-H glycosidic deformation of cellulose (He et al. 2014) was consistent with increasing cellulosic content. Based on FTIR spectroscopic analysis, only the hydrothermal and xylanase treatments of kenaf bark and lignin were found to completely remove hemicellulose. Therefore, the bleaching process was necessary to facilitate complete lignin removal and increase CNF purity.