Optimized isolation and characterization of cellulose for extraction of cellulose nanocrystals from Ensete ventricosum pseudo-stem fibre using a two-stage extraction method

Abstract Alkali treatment followed by alkalinized hydrogen peroxide delignification yielded 73.90% cellulose from Ensete ventricosum pseudo stem fibre, with parameters optimized using response surface methodology. The optimal reaction parameters were 157 min, 73 °C, and 3.8% NaOH concentration. Thermogravimetric analysis (TGA), X-ray diffraction (XRD), Fourier transfer infrared (FTIR), and scanning electron microscopy were used to examine the thermal properties, crystal structure, chemical structure, and morphological structure of isolated cellulose (SEM). Based on the findings, cellulose has a rod-like shape. The XRD results revealed that the crystallinity index of cellulose increased from 65 to 75% when compared with raw E. ventricosum pseudo stem fibre (Ensete fibre). The resultant cellulose demonstrated relatively higher thermal stability than the unprocessed ensete fibre, according to the thermogravimetric examination. When compared to raw ensete fibre, FTIR analysis revealed that cellulose had a modified chemical functional structure, which suggested that alkali and alkalized hydrogen peroxide treatments had altered the chemical structure of cellulose. As a result of the isolated cellulose’s high yield, high crystallinity index, good thermal stability, and morphological structure, cellulose nanocrystals can be extracted. Graphical Abstract HIGHLIGHTS Two step pretreatments were employed to extract cellulose from ensete fibre. Isolated cellulose characterized for it morphological, thermal, functional changes. Cellulose from ensete fibre exhibited good thermal, physical, chemical properties. X-ray diffractometer result depicted that cellulose from ensete fibre has good crystallinity. Cellulose extraction condition was optimized to obtain optimum yield of cellulose.


Introduction
The depletion of petroleum-based resources and a number of environmental issues, including health issues and global warming, have drawn a lot of attention to the creation of ecologically friendly materials [1,2].To solve environmental issues, bio-based materials such as biomass waste, agricultural byproducts, forestry waste, natural fibres, and so on, could be employed as viable substitutes for petroleum-based materials [3].These materials have benefits, including low cost, abundance, biodegradability, renewability, and environmental friendliness [4,5].One of these bio-based materials that potentially takes the place of petroleum-based components is biomass from the Ensete ventricosum plant.
Ensete ventricosum (Enset (Welw.)Cheesman) is a native herbaceous crop with many uses.The Southern, Sidama, South Western, and Oromia regions of Ethiopia are where it is primarily farmed [6].It is a food crop that is only grown in Ethiopia, and little is known about it [7].Bulla, Kocho, and Amicho are three foods made from the crop that are excellent sources of starch and carbohydrates [8,9].Additionally, this plant is an excellent source of fibre and animal feed [10].Enset can be grown in a much larger area to feed more than 100 million people and improve food security in Ethiopia, Kenya, Uganda, and Rwanda, according to recent modelling studies.The plant exhibits adaptability to changing climates and can be cultivated and harvested at any time of the year [7].
The plant is widely cultivated throughout the country to feed a sizable population, and as a result of the massive food production and cultivation, enormous amounts of waste lignocellulose biomass are obtained [11][12][13].The plant's leaves, pseudo-stem (a bundle of fibres), and main stalk of E. ventricosum are the main sources of these residual wastes [14][15][16].According to various studies, the leaf lamina, leaf midribs, pseudo stem, and corm are the four fractional portions that make up the majority of E. ventricosum's dry matter (Figure 1) [13].These four dry matter fractions have different compositions depending on the kind, region, season, etc. of the plant [12,17,18].Different E. ventricosum types have different ratios of these four ingredients on dry bases.Leaf lamina is 6-17%, the pseudo stem is 45-60%, stalk is 4-21%, and corm is 10-30%, according to research findings [13,15,19,20].The pseudo stem fraction is larger than the other fractional sections, as shown in these reports, and it is primarily made up of pulpy parts, long and thin fibres with low extractive, ash, and lignin content (4.4,3.8, and 10.5%, respectively).The pseudo stem fibre (one of the Musa plant's fibres) has a high holocellulose and cellulose content (87.5 and 59.5%, respectively) [11,14,21].Unfortunately, Ethiopian E. ventricosum pseudostem natural fibre resources have not been fully exploited.This natural fibre is mostly made up of two polysaccharides cellulose, hemicellulose, and one aromatic polymer lignin as well as small amounts of moisture, ash, and extractives [15,22].E. ventricosum pseudo-stem fibre's primary polysaccharide, in terms of its chemical composition, is cellulose.
Therefore, there is significant potential in isolating cellulose from E. ventricosum pseudo stem fibre for many applications, including cellulose nanocrystals, polymers, textiles, and so forth.Cellulose is the most important natural product created by living organisms because it forms the structural basis of the cell wall of lignocellulose biomass [23].This polymer, which is essential for preserving the structure of plant cell walls, often comprises of fibrous, water-insoluble, rigid crystalline substance [24].It is one of the major natural polymers in lignocellulose biomass, which is composed of linear β-1, 4 glycosidic chains of glucose molecules that make it a rigid and insoluble crystalline polymer [25,26].Cellulose is made up of microfibrils, which are collections of fibrils.However, each individual fibril has both crystalline and amorphous domains [27].For use in various application areas, non-cellulose components of fibres can be broken down using chemical or mechanical processes to produce cellulose [28].
Cellulose extraction can be seen as one of several effective pre-treatment techniques for the elimination of undesirable components in the development of various products from cellulose in various application areas [29].To be used in a variety of fields, including biomedical, adsorbents, nanocomposite, energy, food, and packaging materials, Nano cellulose required some extraordinary properties, including a unique nanostructure, hydrophilicity, low toxicity, biocompatibility, biodegradability, ease of modification, and high mechanical properties [30][31][32].Various mechanical, chemical, or chemo-mechanical extraction techniques have been published in the past few years to extract cellulose from lignocellulose biomasses [33,34].Different techniques, including hydrothermal extraction, conventional alkali extraction, ambient condition extraction, and two-stage extraction, can be used to extract cellulose [28].One of the most popular methods for biomass to isolate cellulose from and enhance the mechanical and thermal stability of the extracted cellulose for cellulose nanocrystals is alkali solution treatment [35].This study uses the two-stage alkali and hydrogen peroxide delignification extraction process to get enhanced morphology, dimension, chemical composition, hydrogen bonding system, and other properties.
A thorough investigation is needed to develop the use of E. ventricosum pseudo stem fibre, improve its economic potential, and ensure its sustainability.Even though this subject has been the subject of research, it is still vital to examine its potential from those angles in order to make the most of it.The renewability, abundancy, low cost, underutilization, and biodegradability of this natural fibre are of the intriguing qualities that have made it a candidate for usage in military vehicles, textiles, biomedical applications, structural and building construction, automotive, aerospace, and other industries.The goal of this study was to extract cellulose from E. ventricosum pseudo stem natural fibre so that it could be used to isolate cellulose nanocrystals.

Preparation of cellulose
Mechanical treatments and oxidation reaction of the fibre were similar with the one performed in previous research by Abnet Mengesha [11].The removal of large amount of hemicellulose was achieved by alkali treatment.Mechanically treated ground and wax free ensete fibre was treated with solution of sodium hydroxide.The alkali treatment process performed at different reaction condition (sodium hydroxide concentration: 2-5%; reaction time: 120-240 min and temperature: 50-100 °C) and fixed cellulose to sodium hydroxide ratio 1:20 (w/v).Continuous mechanical mixing was employed with speed of 750 rpm.By introducing cold water (4 °C) into the reaction beaker the process was stopped.Double distilled water used to remove stacked sodium hydroxide from the surface of the solid cake.The process was continued until neutral pH was obtained.Then the solid cake was dried for 24 h at 45 °C for the next process [11].

Calculation of cellulose yield
Raw ensete fibre and alkali-treated fibre were studied for their chemical compositions (cellulose, hemicellulose, and lignin).The alkali-treated fibre cellulose content was studied at each reaction point and at the final optimized point.All the experiments were performed three times according to standards, protocols, and previous works [11].

Experimental design
An empirical modelling method used in the creation of experimental models and experiment design using response surface methodology (RSM).The surface plotting and statistical analysis were done with Design-Expert 11.The three independent variables were sodium hydroxide concentration (2, 3.5, and 5%), reaction time (120, 180, and 240 min), and reaction temperature (50, 75, and 100 °C).The yield of cellulose was the outcome, and the selection range for each variable was displayed in Table 1.

Thermogravimetric analysis (TGA)
The thermal behaviour of raw and isolated cellulose from E. ventricosum pseudo stem fibre was performed on a thermal analyzer (HTC_1, Ethiopia) to determine the thermal stability of the samples under a nitrogen atmosphere in a platinum crucible for about 60 min with a flow rate of 25 mL/min.The heating cycle used is a constant heating rate of 20 °C/min from room temperature to 700 °C [11].

Fourier-transformed infrared (FTIR) spectra analysis
Infrared spectroscopy was used to examine the chemical functional groups and structural differences between raw ensete fibre and cellulose.A thermos-scientific (iS50 ABX, Addis Ababa Science and Technology University, Ethiopia) spectrometer was used to gather Fourier transfer infrared (FTIR) spectra with a resolution of 16 cm −1 in the 4,000-400 cm −1 range.To make pastilles, about 5 mg of each sample was first ground into a powder and combined with KBr [5,11].

X-ray diffraction (XRD) analysis
An X-ray diffraction (XRD) examination was used to look into the crystallinity index (%CrI) of raw fibre and cellulose samples.The X-ray diffraction patterns of several powder samples were obtained for this investigation using a diffractometer (XRD-7000 X-RAY DIFFRACTOMETER, Shimadzu Corporation, Japan) that was scanned at room temperature between 2θ with a diffraction angle in the range of 10° to 80° and a step size of 0.02°.The applied current was 30 mA while the accelerating voltage was 40 kV.
Where Iam is the minimum intensity of the amorphous region alone at peak 2θ = 18°, and I200 is the maximum intensity of the peak at 2θ = 22.65°, which corresponds to both the crystalline and amorphous regions.

Scanning electron microscopy (SEM) analysis
Scanning electron microscopy JCM, 6000 plus (Addis Ababa Science and Technology University, Ethiopia) was used to investigate the surface morphology of samples of raw ensete and cellulose at a working distance of 500-10 μm and with an accelerating voltage of 15 kV.Prior to observation, all samples were covered with gold [11].

Model selection and regression analysis
Tables 2-5 are present the sequential model sum of squares, model summary statistics, and analysis of variance for the quadratic model, respectively.The experimental data were analyzed using design-expert software to generate a regression equation, equation, which represented the cellulose yield versus the function of the independent variables.Accordingly, only the quadratic model's p value is 0.0001 in the sum of squares analysis (Table 2), indicating that the model is significant.According to variance analysis, the quadratic model's F value is 2421 (Table 4) and its lack of fit mean square value is 0.007 (Table 4), all of which show that it is appropriate for the experimental design.The ratio of the explained variation to the overall variation, which measures the model's fitness, is generally known as the coefficient of determination R 2 .A high R 2 value implies that the dependent variables in the model are relevant.The quadratic model's R 2 value is 0.998 (Table 3), which shows how well it can account for the system's actual behaviour.
The experimental data were analyzed in Design-expert 11 to generate a regression equation that described the cellulose yield as a function of the independent variables, as shown in Eq. ( 3 Where, the coded variables for the concentration of sodium hydroxide, the reaction temperature, and the reaction time are A, B, and C, respectively.

The analysis of variables interaction effect and optimization
Response surface plots, in general, display the interaction effects of two factors on the response of interest while maintaining the level of other variables (i.e.zero).Elliptical plots show a substantial interaction between the variables, while circular contour plots show an insignificant interaction between the variables.Figure 2 depicts the impact of reaction temperature, sodium hydroxide, and their interaction on cellulose yield.The breakdown of non-cellulose components and the improvement of cellulose yield are both significantly influenced by sodium hydroxide concentration.The alkali treatment of ensete fibre could not produce a larger yield of cellulose at a lower sodium hydroxide concentration.The output of cellulose rises as the concentration reaches a specific point.The best yield may be attained at a sodium hydroxide concentration of (2.6 to 4.4%) and a reaction temperature of 65 to 100 °C.Overreaction is the cause of the decreased cellulose yield at increasing sodium hydroxide concentrations.Cellulose was broken down into its component sugar molecules when an over-reaction with an excess of sodium hydroxide concentration and a higher reaction temperature took place.
The impact of reaction time and sodium hydroxide concentration on cellulose yield is depicted in Figure 3. Reaction time is crucial in the breakdown of non-cellulose  materials.The reaction of the fibre could not take place efficiently in a short amount of time.The yield of cellulose rises as response time reaches a particular point.At 150-190 min of reaction time and a sodium hydroxide concentration of 2.2%-3.9%, the best yield is obtained.The overreaction is the cause of the decline in cellulose yield with increasing reaction times.Cellulose was split into its constituent sugar molecules when the over-reaction took place with too much reaction time.
Figure 4 illustrates how reaction temperature and time affect cellulose production.The breakdown of ensete fibre into its constituent parts, such as cellulose, depends significantly on reaction time.The ensete fibre could not be properly treated with alkali in a short amount of time.The yield of cellulose increases when the reaction time rises to a particular point.The best yield may be attained at reaction temperatures between 65 and 100 °C and reaction times between 150 and 190 min.The overreaction is the cause of the decreased cellulose yield with longer reaction times.Cellulose was disintegrated into its constituent parts when the over-reaction took place with too much reaction time.
The best conditions are 73 °C for the reaction temperature; 157 min for the reaction duration; and 3.8% sodium hydroxide concentration, according to the computer program's forecast (Figure 5).In these circumstances, the yield of cellulose from ensete fibre reaches 73.9%, which is not substantially different from the projected value of 72.91% at a 95% confidence interval (Table 5).

Thermogravimetric analysis
Figure 6 displays the TGA and DTG curves of cellulose and ensete fibre (a) (b).As the figure revealed, the weight loss during the thermal deterioration of samples occurs at three different temperatures.The initial thermal degradation of ensete fibre results in a small weight loss at 192.5 °C, which is followed by a large weight loss between 192.5 and 327.5 °C, and finally a moderate weight loss up to 485 °C.There is a noticeable weight decrease for cellulose between 247.5 and 352.5 °C.Additionally, according to the DTG curves, the maximal weight loss for ensete fibre and cellulose peaks at 350 and 355 °C, respectively.When compared with cotton the TGA patterns of E. ventricosum pseudo stem fibre cellulose is different because of its low cellulose (77.74 and 86.67 for raw and treated fibre, respectively) content and high lignin (6.68 and 1.01 for raw and treated fibre, respectively) content [11].All of the aforementioned findings show that cellulose has greater thermal stability than raw ensete fibre.This is because raw ensete fibre contains more non-cellulosic components than cellulosic components, which encourages heat degradation of the samples [11,36].

Fourier-transformed infrared (FTIR) spectra analysis
Figure 7 displays the FTIR spectra of raw ensete fibre and cellulose.The presence of a stretching vibration of the O-H group, which is related to the intramolecular hydrogen bond of cellulose, results in the characteristic bands of 3277.00-3348cm −1 and 2922.7 cm −1 associated with stretching vibration C-H.The figure clearly demonstrates how, when the fibre is subjected to chemical treatment, the transmittance for this region gradually changes.The transmittance showed an increment and the peak becomes sharper due to alkali treatment and an increase in cellulose content [11,37,38].Lower absorbance intensity ratios are found in the absorption band at 1,634.7 cm −1 , which is associated with the vibration of carbonyl groups and adsorbed water molecules.The opening of the terminal glycopyranose rings or the oxidation of the C-OH groups may be the cause of the occurrence of this peak [39].In ensete fibre, 1327.74 cm −1 corresponds to the CH2 scissoring of cellulose, hemicellulose, and lignin as well as the stretching of the C-O ring of syringyl lignin and the condensed G ring of lignin [22].The peak at 1315.45 cm −1 is associated with the rocking vibration of CH2 in the cellulose alcohol group [40].The peak at 1002 cm −1 illustrates the C-O-C pyranose vibrating stretching ring of cellulose [41].

X-ray diffraction (XRD) analysis
The X-ray diffraction (XRD) patterns of cellulose and ensete fibre are displayed in Figure 8.In accordance with the distinctive diffraction peaks of cellulose I, XRD patterns exhibit diffraction peaks at 2θ = 15.2, 17.4, 22.2, and 34.8, which correspond to the (110), (110), (002), and (004) crystallographic planes [42].At those diffraction angles, sharp peaks are clearly seen, which represent cellulose I type indicating higher crystallinity of microfibers.The results clearly show that cellulose II causes the sharp peaks that are seen at around 2θ (˚) 42, 65, and 78.It was suggested that the polymorphic transition might cause crystallinity to decrease.As clearly observed from the figure, the crystallinity of the sample increases with stepwise chemical pretreatments [43].And it also may be caused by the aluminium sample holder that was used in this study.Ensete fibre and cellulose have a crystallinity index of 44.1 and 62.3%, respectively.The increase in cellulose's crystallinity index has been caused by the reduction and deletion of its non-cellulosic elements and disordered areas [39].Additionally, the crystal lattice in the (002) plane appears to be more perfect than that of an ensete fibre, as seen by the diffraction peak of 22.2 becoming shaper.E. ventricosum's pseudo-stem fibre's crystallinity, abrupt peaks, and cellulose content have all been previously discussed [11].

Scanning electron microscopy (SEM) analysis
Figure 9a,b, respectively, displayed the SEM images of cellulose and ensete fibre.SEM micrographs show that the morphologies of raw ensete fibre and separated cellulose are clearly distinct from one another.Raw fibre image (Figure 9a) reveals a smoother surface, which was a result that was anticipated given the presence of extractives such as wax and pectin [36,44,45].In contrast to raw fibre, processed Ensete fibre had a rough surface with fragments of fibre bundles and an irregular form and size.The primary cause of this is the destruction of non-cellulose components of the fibre by chemical pretreatments, which results in the removal of the protective layer of alpha cellulose [44].The result is comparable with other biomasses [35].

Conclusions
The extraction of cellulose from ensete fibre a plentiful, affordable, and easily accessible material was successful.In the current work, the effects of mechanical treatment, chemical pre-treatment (alkali and bleaching), and its alkali treatment condition on the yield and properties of cellulose have been investigated.These effects affect the functional groups, chemical composition, morphology, thermal properties, and crystallinity of ensete fibre.An extraordinarily high amount of cellulose was produced by the effective and gradual removal of the lignin layer and hemicellulose from the biomass, which makes it easier to extract CNCs afterwards.The amount of cellulose and the reaction conditions have a significant impact on the yield of cellulose.So, attaining a good CNC yield and the right characteristics requires correct reaction conditions for chemical pre-treatment of cellulose.The outcomes for various characterizations show that ensete fibre is a productive, enduring, and regenerative source for the isolation of cellulose nanocrystals.The optimal sodium hydroxide concentration, reaction temperature, and reaction time were found to be 3.8%, 73 °C, and 157 min after optimizing the alkali treatment conditions for cellulose isolation.The substantial potential of the E. ventricosum pseudo stem fibre for the extraction of cellulose nanocrystals is revealed by the high crystallinity index and good thermal stability of the obtained cellulose.

Figure 2 .
Figure 2. response surface (a) and contours (b) depicting the effect of reaction temperature and naoH concentration on the yield of cellulose.

Figure 3 .
Figure 3. response surface (a) and contours (b) depicting the effect of naoH concentration and reaction time on the yield of cellulose.

Figure 4 .
Figure 4. response surface (a) and contours (b) depicting the effect of reaction temperature and reaction time on the yield of cellulose.

Figure 5 .
Figure 5. response surface (a) and contours (b) of the optimized yield of cellulose.

Table 1 .
response to cellulose yield and the Box-Behnken design. )

Table 2 .
analyses of several models using the sum of squares.

Table 3 .
summaries of the model's statistics.

Table 4 .
anoVa for response surface quadratic model.

Table 5 .
summary of independent variables coefficients and standard deviation.