Inuence of Hydrolysis Conditions on Characteristics of Nanocrystalline Cellulose Extracted from Ramie Fibers by Hydrochloric Acid Hydrolysis

Nanocrystalline celluloses (NCCs) were successfully extracted from ramie bers using chemical pretreatments followed by hydrochloric acid hydrolysis. The effects of acid concentration and hydrolysis time on the characteristics of NCCs were investigated in this study. Results showed that the optimal hydrolysis conditions were found to be 6 M hydrochloric acid concentration at 45 °C for 70 min. The obtained NCC had a rod like-shape with an average of 8.07 nm in diameter, 158.51 nm in length, 22.37 in aspect ratio, 89.61% in the crystallinity index, and 5.81 nm in crystallite size. The higher crystallinity and thermal stability were exhibited by NCCs compared to both raw bers and chemically puried cellulose. The hydrolysis time had a signicant effect on crystallinity and thermal stability. The crystallinity index and thermal stability of NCCs were obtained to decrease with increasing hydrolysis time.


Introduction
In recent years, rapid developments and many opportunities for nanotechnology attracted special attention from researchers and industrialists, especially in nanocrystalline cellulose (NCC) [Khalil et al. 2012; Peng et al. 2011;Mishra et al. 2019]. That is due to its superior properties such as high speci c strength, high speci c modulus, high exibility, good dynamic mechanical and unique optical properties, low coe cient of thermal expansion, large speci c surface area, nanoscale dimension, high aspect ratio, low cost, biodegradable, and biocompatibility [Chen et  NCC is a nano-sized material with the structure of rigid rod-like shaped whiskers with a diameter in the range of 1-100 nm and a length in the range of 10-100 nm [Fan et al. 2012;Zainuddin et al. 2017]. NCC is extracted from cellulose that contains amorphous and crystalline regions. Moreover, NCC has a good ordered crystalline structure due to the amorphous zone of cellulose is dissolved during processing [Pirani et al. 2013;Naduparambath et al. 2018]. NCC can be prepared from cellulosic materials such as bacteria, algae, animal (tunicate), plants, and some waste material [Jasmani and Adnan 2017; Ilyas et al.
Extracting NCC from lignocellulose materials can be done through chemical pretreatment (cellulose purifying) to remove amorphous regions (wax, lignin, pectin, and hemicellulose) followed by acid hydrolysis. Some commonly used mineral acids include phosphoric acid, sulfuric acid, hydrochloric acid, and their mixtures [Jasmani and Adnan 2017; Liu et al. 2014]. Cellulose hydrolysis aims to eliminate amorphous cellulose areas to obtain cellulose crystalline regions [Oun and Rhim 2016]. The most popular method is acid hydrolysis by sulfuric acid [Rosa et al. 2010] where this method causes the extracted NCC to have a unique surface property of SO 3 − that makes NCC have a good dispersion in water solvent [Yu et al. 2013]. However, the phenomenon will decrease the thermal stability of NCC [Corrêa et al. 2010]. On the other hand, the extraction of NCC through hydrochloric acid hydrolysis can produce a minimal surface charge in extracted NCCs [Oun and Rhim 2015], and the thermal stability of NCC is higher than NCC extracted by sulfuric acid [Corrêa et al. 2010]. Therefore, hydrochloric acid was chosen to obtain NCC with good thermal stability in this work. To date, a study on isolation and characterization of NCC from ramie bers by hydrochloric acid hydrolysis has not yet been reported.

Isolation of nanocrystalline cellulose
Before isolation of nanocrystalline cellulose, the ramie bers were chemically pretreated to obtain chemically puri ed cellulose (CPC) through chemical pretreatments (de-waxing, bleaching, and alkali treatments) as described in our previous study [Listyanda et al. 2020]. The obtained CPC was then isolated for producing NCC using hydrochloric acid hydrolysis as described in the literatures [Yu et al. 2013;Kumar et al. 2014]. Brie y, a certain amount of distilled water was poured in a beaker glass and dropped by hydrochloric acid (37.8%) that was placed on the ice water bath under the continuous stirring on the magnetic stirrer until the nal HCl concentration was 6 M at 20 ml acid solution. At the end of acid addition, the ice bath was removed, and the acid solution was preheated until 45 °C. After that, 1 g of CPC was added and kept for 70 min. In this work, the effects of two hydrolysis conditions (acid concentration and reaction time) on the characteristics of NCCs were investigated. A series of NCC suspensions were prepared using various hydrochloric acid concentrations (6,8, and 12 M) at 45 °C for 70 min and they were then denoted as NCC-6, NCC-8, and NCC-12, respectively. In addition, NCCs were also prepared with 6M HCl at 45 °C using three hydrolysis times which are 70, 125, and 180 min, denoted as NCC-70, NCC-125, and NCC-180, respectively.
After hydrolysis processes were completed, the suspensions were quenched into the chilled distilled water (temperature ± 5 °C) with a ratio of suspensions to chilled water 1:20 (v/v) to end the hydrolysis process. The suspensions were then centrifuged at 4,000 rpm for 15 min to release the acidic solution. The supernatant was then removed and replaced by distilled water followed by centrifugation at 4,000 rpm for 15 min to optimize the removal of the acidic solution. After centrifugation, the NCC sediment was collected and then re-suspended by distilled water and dialyzed against distilled water until neutral (pH 7). After that, the NCC suspensions were ultrasonicated for 2 min with a 50% amplitude to produce the homogeneous NCC suspensions. Furthermore, some parts of NCC suspensions were collected and others were freeze-dried for characterizations.

Chemical composition
The chemical composition analysis of the bers was performed to determine the content of the three main components of the bers, namely: α-cellulose, hemicellulose, and lignin. This was carried out on the raw bers and CPC. The chemical composition analysis was performed following SNI 0492: 2008 standard. .

Fourier Transform Infra-Red (FT-IR)
FT-IR spectra were recorded using the IRPrestige21 machine in the range of 4000 to 500 cm-1 were used to analyze the functional group of samples. Two milligrams of samples were mixed with 200 mg KBr then mixed using a vibrating mill for 30 seconds to make a homogeneous mixture, the mixture was then compacted under vacuum condition to form to pellets and then put into the FTIR machine to obtain its spectra.

X-Ray Diffraction (XRD)
XRD was used to calculate the crystallinity index and the crystallite size of the samples. The XRD patterns were analyzed using Empyrean (Malvern PANanalytical) in range of 2θ = 10 to 45° using Niltered CuKα ray radiation (λ = 0.154 Å) operated at 40 kV and 30 mA in room temperature. Furthermore, those patterns were used to calculate the crystallinity index and crystallite size of the samples. The crystallinity index of the samples was determined by using the Segal method [Segal et al. 1959] that can be seen in Eq. 1: where I 200 is the peak intensity of the diffraction at (200) peak for cellulose I at around 2θ = 22° and I am is the intensity related to the amorphous scattering at around 2θ = 18°. The crystallite size of the samples was determined by using the Scherrer equation [31] following Eq. 2 as follows: where t is the crystallite size, K is the Scherrer constant (0,91), λ is the X-ray wavelength (λ = 1.54060 Å), β 1/2 is the full width at the half maximum (FWHM) of the XRD peak in radian and θ is the Bragg's angle.

Scanning Electron Microscopy (SEM)
SEM was used to observe the effect of chemical pretreatment on the surface morphology of the bers. This observation was carried out on the raw bers and CPC using ZEISS type EVO MA 10 -SEM with an accelerating voltage of 5 kV. Before observing, the samples were coated with palladium by a sputtering method to create a conductive thin layer on the surface of samples.

Transmission Electron Microscopy (TEM)
The morphology and dimension of NCC were observed using TEM (JEOL type JEM-1400 Electron Microscope) at an accelerating voltage of 120-200 kV in several different magni cations. Firstly, the NCC suspension was homogenized using an ultrasonic homogenizer for 60 seconds. A drop of the NCC solution was then deposited on a carbon-coated copper grid which was dried and observed for imaging. The diameter and length of NCCs were measured by an image processing analysis program, Image J, using the TEM images.

Thermal Analysis
The thermal stability was characterized by using thermogravimetry analysis (TGA) (LINSEIS machine Type TA PT 1600). The experiments were performed at a heating rate of 10 ˚C/min from temperatures of 30 to 600 ˚C under the nitrogen atmosphere at a ow rate of 4 liters/hour. Besides, the derivative thermogravimetry (DTG) was obtained from TGA data by calculating using a central nite difference method that can be seen in Eq. 3: where w t+Δt is the weight of the sample at time t + Δt, wt-Δt is the weight of sample at time t -Δt, and Δt is the time interval for reading residual sample weight [Oun and Rhim 2015].

Chemical composition
From the chemical composition analysis, it was found that the raw bers consist of 72.68% α-cellulose, 13.70% hemicellulose, 0.38% lignin, and 13.24% others (water, wax, and pectin). This suggests that ramie ber has high cellulose content and has the potential to be used as a renewable source of cellulose.
Therefore, this was the reason for choosing of ramie bers as a source of cellulose for the preparation of nanocrystalline cellulose. Furthermore, the CPC compose of 92.30% α-cellulose, 7.18% hemicellulose, 0% lignin, and 0.52% others. This means that the chemical puri cation signi cantly raised the α-cellulose and reduced the contents of amorphous regions such as hemicellulose, lignin, wax, and pectin. It is wellknown that the highest polymerization degree and the most stable among all classes of cellulose is α- cm ¹ (C-O-C glycoside bonds asymmetrical stretching), 1111 cm − 1 (C-OH stretching), 897 cm − 1 (βglycosidic bonds bending) presented to the cellulose I characteristics was more obvious for NCC. This reveals the NCC had the highest cellulose content among all samples. For NCC, the peak at 1058 cm ¹ (C-O-C pyranose ring vibration) displayed sharper compared to others indicating the higher cellulose content for NCC [Ngwabebhoh et al. 2018]. It can be concluded that the chemical puri cation and hydrochloric acid hydrolysis removed some parts of amorphous components such as lignin, hemicellulose, and wax of bers. FTIR results showed the cellulose Iβ structure and the hydrochloric acid hydrolysis resulted in higher cellulose content than both raw ber and CPC.  Fig. 2, it can be concluded that the acid concentration and hydrolysis time did not in uence the crystal structure of the raw ber as indicated with no different peaks for all samples. These results are consistent with the FT-IR spectra as shown in Fig. 1. Similar ndings were also demonstrated by previous researchers [Ngwabebhoh et al. 2018] where the crystal structure of cellulose was not in uenced by the acid concentration, temperature, and reaction times during hydrolysis.

X-Ray Diffraction (XRD)
From Fig. 2, the crystallinity index of the samples is then calculated using the Segal equation [Segal et al. 1959] and the results are summarized in Table 1. The crystallinity index values of raw ber and CPC are 79.75 and 86.68%, respectively. The slight increase was attributed to the removal of hemicellulose and lignin during the chemical puri cation process of cellulose ]. The effect of hydrochloric acid concentration on the crystallinity index of NCCs as indicated by NCC-6, NCC-8, and NCC-12 is also presented in Table 1. It can be seen that NCCs had slightly higher values of the crystallinity index compared to both raw ber and CPC. The further removal of amorphous regions during acid   Table 3 that the crystallite size slightly decreased with increasing the acid concentration. A similar trend could also be observed when the reaction time increased where the crystallite size was also reduced. A higher reduction in the crystallite size was exhibited by the increased reaction time compared to acid concentration. This indicates the reaction time had a higher effect than the acid concentration in this study. The dissolution of both the amorphous areas and part of crystalline domains during hydrolysis might be responsible for the reduced crystallite size with increasing the acid concentration and reaction time [Samir et al. 2004]. Similar ndings were also demonstrated by previous researchers [Lei et al. 2019] where the crystallite size of NCCs decreased by increasing the sulfuric acid concentration.  Figure 3 displays the SEM micrographs of raw bers and chemically puri ed cellulose (CPC). The chemical treatments were done through de-waxing, bleaching, and alkali treatments. The raw bers were consists of the individual bers linked together as bundles by cement materials (Fig. 3a). The raw ber had a rough surface layer with the presence of high amorphous components such as hemicellulose, lignin, wax, pectin, and oil [Fardioui et al. 2016]. Furthermore, de-waxing, bleaching, and alkali treatments allowed to remove the amorphous components such as hemicellulose and lignin present in the untreated ramie bers, resulting in de brillation of raw bers to the cellulose bers (Fig. 3b). Therefore, the average diameter of bers was decreased from 60 µm for the raw bers to 45 µm for CPC. Similar ndings were also demonstrated by previous researchers where the bleaching and alkali treatments reduced the diameter of raw bers [Rosli et al. 2013;Fardioui et al. 2016]. Visually, the raw bers had a yellowish color whereas CPC was a pure white. This con rms the success of cellulose puri cation through chemical treatments. Furthermore, the morphology of obtained NCCs was observed by using the TEM examination and discussed later. The diameter and length distribution of NCCs particles measured from TEM images are shown in Fig. 5. The dimensions of NCCs were analyzed by measuring particle length (L), diameter (D), aspect ratio (L/D), and the results are summarized in Table 2. The diameter and length of NCCs were calculated by an image processing analysis program, Image J using TEM images.  Table 2, it can be seen that the average dimensions in diameter and length of NCC decreased with increasing both acid concentration and reaction times. In other words, both acid concentration and reaction time had a signi cant effect on the dimensions of NCC. Furthermore, it can be observed from Table 2 that the average aspect ratios of NCC-6, NCC-8, and NCC-180 were 20.67, 22.29, and 23.81, respectively. This indicates that the aspect ratio of NCC particles was enhanced slightly as both acid concentration and reaction time increased. In addition to hydrolysis conditions, the dimensions and aspect ratio of NCC particles were strongly in uenced by sources of cellulose [Habibi et al. 2010]. The aspect ratio of obtained NCCs in this study was still lower compared to that of others presented on NCCs from other cellulosic sources, such as sisal (60)

Thermal Stability
The thermal stability of raw ber, CPC, and NCCs produced under different hydrolysis conditions was characterized using thermogravimetric analysis (TGA). The resulting TGA thermograms and derivate thermogravimetric analysis (DTG) curves are shown in Figs. 6 and 7, respectively and the results are presented in Table 3. Figure 6a displays the TGA curves of raw ber, CPC, and NCCs prepared under different acid concentrations at 45 °C for 70 min. From Fig. 6a, it can be observed that all samples exhibited a relatively small weight loss in the temperature range of 25-150 °C which was mainly due to the evaporation and removal of moisture absorbed in the sample. In addition, the weight loss at this temperature might also be attributed to compounds with low molecular weight or volatile material remaining and sticking to the samples during the cellulose puri cation process [Ishak et al. 2012]. From Fig. 6, it can be seen that the moisture of raw ber was completely evaporated at a higher temperature (around 150 °C) compared to CPC and NCCs. This was associated with the higher moisture content of raw ber resulting in higher weight loss (15%) compared to CPC and NCCs [Ilyas et al. 2018]. Furthermore, the higher moisture content in the raw ber was due to its high hemicellulose content (13.7%) where hemicellulose is an amorphous component with very high water absorption property [Ishak et al. 2012]. From Fig. 6, it can be also observed that all NCCs and CPC exhibited lower moisture content compared to raw ber due to the elimination of amorphous regions (hemicellulose, lignin) and dehydration of cellulose bers during acid hydrolysis. Figure 7 shows the DTG curves of raw bers, CPC, and NCCs prepared under different hydrolysis conditions. The onset temperature of major decomposition of raw ber, CPC, and NCCs is presented in Table 3. The major decomposition of all samples started around 158-254 °C. From Table 3, it can be seen that the onset temperatures of raw ber, CPC, and NCCs were 158, 242, and 233-254 °C. The onset temperature of CPC was higher than the raw ber due to the lower amorphous region of the CPC resulting in higher thermal stability [Sonia and Dasan 2013]. Furthermore, it can be also seen from Table 3 that both CPC and NCCs had higher thermal stability than the raw bers. This was again associated with the removal of amorphous regions (hemicellulose, lignin) during the chemical puri cation process and acid hydrolysis as con rmed by the results of FT-IR and XRD previously. The more crystalline component of NCCs compared to both raw ber and CPC might be believed to be responsible for the higher thermal stability of NCCs [Mohamed et al. 2017]. In addition, the lower decomposition temperature of raw ber was associated with the low degradation temperature of hemicellulose, lignin, and pectin [Chen et al. 2011;Nasri-Nasrabadi et al. 2014]. Furthermore, the onset temperature of NCC prepared under acid concentrations of 6, 8, and 12 are 254, 253, and 252 °C, respectively. No signi cant difference in the onset temperatures suggests that the acid concentration did not affect the thermal stability. This is consistent with the results of the crystallinity index as shown in Table 3. The effect of reaction time during hydrolysis on the thermal stability was characterized by the values of the onset temperature of major decomposition. From Table 3, it can be found that the onset temperature of NCCs prepared under different reaction times of 70, 125, and 180 min is 254, 233, and 235 °C, respectively. It is interesting to note that the onset temperature was decreased drastically with increasing the reaction time. This indicates that the thermal stability of NCC reduced with an increase in reaction time and the best thermal stability was obtained for the hydrolysis time of 70 min. Longer reaction time during hydrolysis led to increase in the number of short cellulose chains due to the highly speci c surface are and free end chains of nanocrystalline cellulose and then resulted in the reduced thermal stability [Nasri-Nasrabadi et al.

2012
]. These end chains began to degrade at low temperatures [Wang et al. 2007]. Moreover, some damage on the crystal structure due to longer reaction time might be responsible for the reduced thermal stability of NCC [Jiang et al. 2017]. It can also be seen from Table 3 that the nal residue left 600 °C of the NCCs was higher than both the raw ber and CPC. This was attributed to a higher level of crystallinity of NCCs compared to both raw ber and CPC. In addition, the higher solid residue observed for NCCs was probably ascribed to the changes of cellulose structure during moisture evaporation at around 100 °C, which reduced the oxygen content in the NCCs, and thus producing carbon structures with higher thermal stability [Kim et al. 2001]. The nal residues are affected by several factors including the chemical structure, crystallinity, source of cellulose materials, and hydrolysis conditions for isolation of nanocrystalline cellulose [Oun and Rhim 2016]. From Table 3, it was also found that the maximum degradation temperature (T max ) of raw ber, CPC, and NCCs were 346, 337, and 328-357 °C, respectively.
The lower T max exhibited by CPC compared to the raw ber was ascribed to the lower amorphous components (hemicellulose, lignin) of CPC due to chemical pretreatments. Further removal of amorphous regions during acid hydrolysis might be believed to responsible for the reduced T max of NCCs compared to both raw ber and CPC. From Table 3, it can also be seen that the T max of NCCs was unchanged with increasing the acid concentration. It was found that the acid concentration did not in uence the T max .
Furthermore, there was a signi cant effect from the reaction time where T max of NCCs was increased with increasing the hydrolysis time.

Conclusion
Nanocrystalline celluloses (NCCs) were successfully extracted from CPC using hydrochloric acid with varying the acid concentration and reaction time at a constant temperature of 45 °C. The chemical structure of cellulose did not change during both chemical treatments and acid hydrolysis processes as con rmed by FT-IR and XRD results. The acid concentration did not affect the crystallinity and thermal stability whereas the reaction time in uenced strongly the crystallinity and thermal stability of NCCs. The higher thermal stability was exhibited by NCCs compared to both raw ber and CPC. The crystallinity index and thermal stability of NCCs were reduced with an increase in the hydrolysis times. The optimal hydrolysis conditions were found to be the reaction time of 70 min with 6 M hydrochloric acid at 45 °C, resulting in NCC with a rod-like shape with high crystallinity index (89.61%), the crystallite size (5.81 nm), the average diameter (8.07) and average length (158.51 nm). The obtained NCCs had great potential to be used as reinforcing agents in nanocomposites.  This is a list of supplementary les associated with this preprint. Click to download.