89 Optimization of alkali concentration in the pretreatment of sugarcane bagasse for ethanol production S. M. Asaduzzaman Sujan1,3*, M. Hossain2, M. Nashir Uddin4 and A. N. M. Fakhruddin1 1Department of Environmental Science, Jahangirnagar University, Savar, Dhaka-1342, Bangladesh 2Institute of Fuel Research and Development, Bangladesh Council of Scientific and Industrial Research, Savar, Dhaka-1350, Bangladesh 3Leather Research Institute, Bangladesh Council of Scientific and Industrial Research, Dhaka-1205, Bangladesh 4Pulp and Paper Research Division, Bangladesh Council of Scientific and Industrial Research, Dhaka-1205, Bangladesh Abstract This study was aimed for the investigation of the effect of pretreatment procedure of alkaline, based on the chemical arrangement, surface morphology, structural composition and enzymatic assimilation of sugarcane bagasse for sugars and ethanol production. Alkali pretreatment (0 to 8% w/v of NaOH) assists to reduce the lignin portion (from 19.57±0.03% to 9.91±0.02%) and increase the cellulose content of the treated SB (from 34.66±0.05% to 63.58±0.05%) simultaneously. The optimal conditions for alkali pretreat- ment were 8% NaOH charge at 100oC for 90 min. Enzymatic digestibility of alkali treated SB was significantly improved and hydrolysis yield reached to 89.59% glucose and 61.23% xylose at an prime level using Trichoderma viridae. Further hydrolysate of 8% (w/v) alkali treated SB sample was fermented by Saccharomyces cerevisiae to convert sugar into ethanol and yield was 16.81±0.32% in 24 h. Alkali pretreatment was found to be a treatment of choice for cellulose hydrolysis in SB and subsequent sugar acquired for the production of ethanol during fermentation. Keywords: Sugarcane bagasse; Alkali treatment; Time; Temperature; Bioethanol; XRD; SEM *Corresponding author’s e-mail: asad2306@gmail.com Available online at www.banglajol.info Bangladesh J. Sci. Ind. Res. 58(2), 89-98, 2023 Introduction Due to a possible source of renewable energy and a reduction in Green house gases (GHG) emissions, recently emphasize on exclusive products using lignocellulosic materials has increased. (Den et al. 2018; Janker-Obermeier et al. 2012). In comparison to first generation biofuel, second generation biofuel made from lignocellulosic sources has more energet- ic, economic, and environmental benefits (Hirani et al. 2018). Mostly three biological polymers- cellulose, lignin and hemicellulose comprise the lignocellulosic biomass in which supramolecular assocition of cellulose and its combination with lignin and hemicellulose provide physical and chemical barriers in plant tissue. In general the purposes of pretreatment are (1) enlargement of the available surface area and break- down the cellulose crystal (2) partial depolymerization of cellulose, (3) solubolize lignin and/or hemicellulose (4) modification of enzyme accessibility to improve digestibility of cellulose (Zhao et al. 2012). The process design that has a clear impact on the cellulose, hemicellulose and lignin fragments should be consider solemnly while selecting the pretreatment techniques and constrains (Alvira et al. 2010). Agricultural remains in Bangladesh could be the potential source of bioethanol production. Sugarcane bagasse is loosely bonded anatomical metric, composed of vasculer bundles surrounded by non-fibrous parenchymatic cell (Jahan et al. 2009). Usually it is available as agricultural residue and byproduct from sugar mill indus- tries. In the year of 2015-16, there was 2,95,162 metric ton sugarcane bagasse was obtained from sugar mills of Bangla- desh (http://www.bsfic.gov.bd). To enrich cellulose content in biomass for enzymatic saccarification, current research has focused on various pretreatment processes. Currently, scientists have gone through a lot of studies on different pretreatment techniques to remove the compact and rigid composition by open up the cellulosic structure (Alvira et al. 2010). Pretreatment technologies are commonly abandoned to reduce the structural barriers and boost cellulose availabili- ty based on not only chemicals such as acid, alkali, oxidant, etc. but also several treatment settings. Alkali treatment is one of the most widespread and economical methods used for surface modification of lignocellulosic biomass. In the field of biorefinary, alkali pretreatment is intensively employed to develop the cellulosic materials both mechan- ically as well as chemically and such properties include tensile strength, dyeability, stability of dimension and reactivity (Wu et al. 2011). In alkali pretreatment biomass is treated under moderate reaction conditions ensuring inexpensiveness, inflated recycling possibilities of water and chemical agents (McIntosh and Vancov, 2010). Usually alkali treatment is the most effective pretreatment for agricultural residues, herbaceous crops and hardwood containing low lignin content in the compariosn of softwood containing high lignin content (Agbor et al. 2011; Canilha et al. 2012). This research aims is to investigate the alkaline pretreatment repercussions based on the chemical structure, surface morphology, structure and enzymatic digestion of sugarcane bagasse for sugars and ethanol production. Materials and methods Sugarcane bagasse (SB) was taken from street juice vendor. Warm tap water was used for washing purpose to eliminate the residual free sugars once again. This washing and pressing step was then repeated for three times. After that drying was done in an oven at 65oC for 16 h. Crushing of dried bagasse was done successively using a locally made crusher and sieved (Retsch, D-42759, HAAN, Germany) to have the particle size of 20-40 mesh on average. Before usage, the 20-40 mesh bagasse was kept at room temperature in an airtight plastic container (Sujan et al. 2018). Moisture and Ash content Moisture content of raw materials was measured using the ASTM D 4442-07 procedure. In a glass crucible, 2 grams of pretreatment sample containing 20-40 mesh were dried in oven at 105±2°C. The moisture content was expressed in percent wet basis and weight measurements were taken every 3 h. A muffle furnace was used for burning the Oven-dried samples at 575±25°C for ash content determination follow- ing the ASTM Standard E 1755-01. Volatile matter The volatile matter analysis was carried out in accordance with ASTM Standard D 271-48. Four grams of raw materials were heated in a furnace at 950±20°C for seven minutes. The weight loss, excluding the weight of moisture pushed off at 105°C, is then used to calculate volatile matter. Fixed carbon The difference between 100 and the total of volatile matter, moisture, and ash content was used to compute the fixed carbon percentage. Chemical analysis The technical association of the pulp and paper industry (TAPPI) method detects α-cellulose (T 203 cm-99), pentosan (T 223 cm-01), klason lignin (T211 om-83), and acid soluble lignin (T UM 250) on a dry basis. Ultimate analysis Ultimate analysis of samples was done by following the procedure ASTM Standard D 5291-02. Organic elemental analyzer (Flash 2000, Thermo Scientific, USA) was used with a specific condition (Reactor temp. 900°C, He: 250 kPa, O2: 250 kPa, TCD). Low concentration alkali pretreatment Different alkali concentration (0%, 2%, 4%, 6% and 8%), time (60, 90 and 120 min) and temperature (80, 100 and 120oC) were applied on raw material to check alkali effect on SB shown in Table 3. A portion of the ground bagasse 20-40 mesh (10g) was taken into plastic zipper bag (temperature 80oC, figure 2a and 2b) and stainless-steel reactor (tempera- ture 100oC and 120oC, figure 2c and 2d). Different alkali concentration was applied during the pretreatment process (alkali to SB ratio ranging between 1:12). The mixture was heated at a particular temperature such as in a water bath (80oC) and oil bath (100oC and 120oC) for a desired length of time. In the course of pretreatment process the sample was manually mixed 2-3 times to attain proper alkali treatment. The treated SB was placed into a polyester bag to remove excess alkali water by pressing it. After that it was vigorously washed with tap water (repeated for five times) to remove the remaining alkali. Finally the SB was dried in an oven at 65oC for 72 h and stored in a close container at room temperature for further experiment. Regression model Regression model has been developed for prediction of α-cel- lulose, pentosan and lignin in SB. The general form of the model is: y=α+α1x1 + α2x2+........................+αnxn + ε.......................(1) where α is the constant term. αi are the coefficient of variables xi. ε is the random error term which is minimized with Simple Least Squares Regression (SLSR). Regression coefficients of the independent variables namely alkali concentration, temperature and time are estimated by SLSR method for developing regression model to predict α-cellulose, pentosan and lignin. Efficiencies of these models are expressed by coefficient of multiple determination (R2) and Adjusted R2. 2.9 Separate hydrolysis and fermentation (SHF) The Hydrolysis experiment took place in 100 ml conical flux 10ml enzyme solution with 200 mg (2% dry wt.) in citrate buffer (0.05 M, pH 5.0) at 50°C for 48 h. In this case Trichoderma viride was used for hydrolysis. Hydrolysate was then heated for 15 min in a boiling water bath and centrifugation was done to remove solid particles. The supernatant was used for analysis of released sugars as described by Jamal et al. (2011). During fermentation process according to Firoz et al. (2012), 100 ml media was prepared and 0.5 g of commercial yeast Saccharomyces cerevisiae was used as inoculum which showed good performance to converts sugar into bioethanol. This inoculated media mixture was poured in a suitable glass- ware and was kept in a shaking incubator for 48 h. 10 ml of this medium was then added into the flask and it was properly covered with aluminum foil. Then it was placed in the incubator at 30oC for 24 h, 48 h, 72 h and 96 h for fermenta- tion of sugars to bio-ethanol according to Sujan et al. (2018). Samples from hydrolysis and fermentation were performed by HPLC. High performance liquid chromatography (HPLC) In characterize part, concentration of Sugars and ethanol were determined by HPLC (Ultimate 3000, Thermo Scien- tific, USA) method using Hyper Rez XP carbohydrate H+ 8 µm column (100×7.7 mm) equipped with a Refractive Index (Shodex RI-101) detector. The mobile phase was degassed with deionized water with a flow rate of 0.7 ml/min and column temperature was maintained at 70°C. It is possible to measure the total sugar concentration in the hydrolysis liquid fraction by comparing its peak area detected by HPLC with peak area of 1% standard sugar which consists of two sugars namely glucose and xylose (Sujan et al. 2018). The same column which is specialized for fermentation broth analysis is used for ethanol detec- tion. The kinetic parameters of ethanol fermentation were determined as follows (Islam et al. 2019): Crystallinity measurement X-ray diffraction (XRD) was used to determine the crystalline structure of the SB samples using a diffrac- tometer (GBC XRD) and filtered copper K radiation (λ = 0.1542 nm) by a monochromator at 35.50 kV voltage and 28 mA current, with a speed of about 2o/min and scan- ning in the range of 10 - 80ºC. The crystallinity index (CrI) was obtained from the ratio between the intensity of the 002 peak (I002, 2θ = 22.5) and the minimum dip (Iam, 2 θ = 18.5) according to the following equation (Roberta et al., 2012): CrI (%) = [(I002 - Iam)/I002] ×100 ...................................... (2) where I002 is the highest peak intensity of plane 002 and Iam is related to the amorphous structure. In present study, the average crystallite sizes were deter- mined from the Scherrer equation by using the diffraction pattern obtained from the 002 (hkl) lattice planes of cellulose samples D(hkl) = [(Kλ / B(hkl) cos2θ] ............................................... (3) Where D(hkl) (Crystallite size), K (Scherrer constant, 0.84), λ (X-ray wavelength, 0.154nm), B(hkl)(Full width half maxi- mum of the measured hkl reflection), and 2θ (Corresponding Bragg angle). Scanning electron microscopy (SEM) analysis In this research, SEM (ZEISS EVO 18 SEM) was used to detect the change of pretreated bagasse fibers. SEM images were taken of different pretreated bagasse samples with acceleration voltage of 2.0 KV. Results and discussion Proximate analysis Proximate analysis of SB sample (20-40 mesh) are presented in Table I. Primarily this analysis usually evaluate the fuel characteristics of raw materials. According to Sun et al., 2009, higher moisture and ashcontent in samples lessen the heating value. Ultimate analysis Ultimate analysis denotes the elemental configuration of SB such as carbon, hydrogen, oxygen, nitrogen and sulfur which are shown in Table II. This examination helps to measure the percentage of carbon and hydrogen content in biomass that is responsible to determine the amount of air is required for complete combustion, composition of combustion gases and heat is generated by it (Poddar et al. 2014). Chemical properties Raw SB contained 34.66±12% α-cellulose, 22.43±08% pentosan and 19.57±06% klason lignin in which 1.75±04% acid soluble lignin (dry basis) was detected by technical association of the pulp and paper industry (TAPPI) method. The chemical composition of SB was determined by acid hydrolysis and it was calculatedby HPLC method as 45.35% glucose and 30.64% xylose (Sujan et al., 2018). α-cellulose yield Bagasse is mainly composed of cellulose, hemicellulose and lignin. Besides these there are some extractives such as ash, wax, gum, pectin etc. During alkali pretreatment usually most of the extractives are removed with the increasing of alkali concentration. Pretreatment of SB with different alkali concentration based on raw material (0%, 2%, 4%, 6% and 8%), time (60, 90 and 120 min) and temperature (80, 100 and 120oC) are shown in Table III. Consequences of each independent experiment varying with alkali concentration, temperature and time on the α-cellulose yield were analyzed using MATLAB software. Apparently, it was observed that α-cellulose yield was considerably increased with changes of alkali concentration ranging from 0% to 8% (34.66% to 63.58%). But no noticeable variation was observed in temperature-time alteration during pretreatment of SB. Based on cellulose percentage obtained in treated bagasse, the optimum conditions for pretreatment reaction were selected as alkali concentration 8%, time 90 min and temperature 100oC. Although cellulose content in treated SB was slowly increased with the increase of alkali concentration 12% attime 90 min and temperature 100oC but as a consequence of huge chemical consumption, recovery problem, chance of losses cellulose and hemicellulose, alkali concentrations for pretreatment above 8% was not considered as ideal concentration. Effect of alkali concentration (AC), cooking time and temperature (temp) charge on α-cellulose yield as well as their statistical significance on the basis of F-test number are presented in regression equations (4). As shown in equations, cooking time at the maximum temperature had no significant effect on α-cellulose yield followed by alkali concentration charge. Effect of temperature on α-cellulose yield was less in employed cooking conditions. For α-cellulose yield: α-cellulose yield = 82.05-0.151×temp-1.56×time-0.69×AC (R2=0.89, adjusted R2=0.87) ............................................. (4) For Pentosan: Pentosan = 13.58+0.086 × tem p + 0.006 × time+0.445 × AC (R2=0.67, adjusted R2=0.63) ............................................. (5) For Lignin: Lignin = 27.662-0.041×temp-0.001×time-01.242×AC (R2=0.93, adjusted R2=0.92) ............................................. (6) For predicting α-cellulose, percentage of pentosan and lignin, the most influential factor was alkali concentration and then cooking temperature for α-cellulose yield, which exhibited an almost linear dependence on both operational variables. The coefficient of determinations is good for α-cel- lulose yield and lignin percentage which hovers around 90 percent, although the figure is moderate more than 60 percent for pentosan. All these three models are significant (p<0.05) at 5% level of significance. In order to perceive the impact of alkali concentration and temperature on these three parameters, three-dimensional (3D) response surface plots were created by plotting the response (α-cellulose yield) pentosan and lignin on the Z-axis versus the most influential one independent variable alkali concentration and temperature as shown in Fig. 3 (a), (b) and (c). XRD analysis In 19th century the cellulose crystalline structure has been discovered and later it was verified by X-ray crystallography (Meyer and Misch, 1937; Wilkie, 1961). The crystallinity index (CrI) of non-woody biomass, such as grasses and agricultural residues, varies depending on the type of biomass. Generally, non-woody biomass has a lower crystallinity index than woody biomass, typically ranging from 20-40%. The crystallinity index of non-woody biomass affects its digestibility and energy production potential.Recently, researchers are being paid more atten- tion on cellulose index because of its potential use in bioen- ergy production. Since then several different models of cellulose index have been proposed. The most popular two-phase cellulose model describes cellulose chains as containing both crystalline (ordered) and amorphous (less ordered) region (Park et al., 2010). Alkali treatment of bagasse has been found to increase the crystallinity index of the bagasse. Alkali treatment breaks down hemicellulose, making the cellulose molecules more ordered and crystalline, resulting in a higher crystallinity index. Alkali treatment also increases the digestibility of the bagasse, making it easier to break down and therefore more suitable for bioenergy production. X-ray diffraction spectra of the SB and alkali treated SB were presented in Figure 4. It was observed that the intensity of 101 and 002 peaks were gradually increased. The relative amount of crystalline cellu- lose (CrI) in the total solid were calculated based on the equation (2) and obtained 18.68%, 24.80%, 27.23%, 32.69%, 36.45% and 39.19% for raw SB, 0%, 2%, 4%, 6% and 8% alkali treated SB samples respectively. The CrI of alkali treated SB samples (not cellulose crystallinity) were intense- ly influenced by the composition of the samples. In case of lignocellulosic biomass examples, cellulose CrI measured the relative amount of crystalline cellulose in the total solid. Therefore, amorphous part of lignin and hemicellulose in biomass specimens were partially removed with the delig- nification process as a result the proportion of α-cellulose was increased and hence CrI would be increased gradually. This interpretation could be proved by the fact that alkali-treated SB had higher CrI than raw SB (Zhao et al. 2010) and the portion of cellulose in the treated SB was also increased gradually. According to the equation 3, the average sizes of crystallite obtained were 3.28, 3.51, 3.63, 3.78, 3.90 and 4.06 nm for raw SB, 0%, 2%, 4%, 6% and 8% alkali treated SB samples respectively. The experimental data revealed that during delignification, the size of crystallite was increased. SEM analysis SEM is one of the most commanding tools widely used to inves- tigate the lignocellulosic biomass surface (Amiri and Karimi, 2015). SEM is usually employed for surface characterization, morphology and inspection of microstructure. In respect of biomass example through SEM images we can compare the untreated and pretreated models which may lead to different insight into the biomass (Karimi and Taherzadeh, 2016). SEM provides two-dimensional images of raw SB and alkali treated SB, which were taken and compared to the outcome of NaOH treatment. All testers were coated with carbon tape and magnification of 500x was used. Figure 4 shows the wall of raw SB (Figure 6a) was intact where the alkali treated SB (Figures 6c-6f), the cell wall was ruptured or splitted and hence packing of the fibers were partially loosened (Firoz et al. 2012). Sugar and ethanol yield Enzymatic digestibility is the ability of enzymes to break down molecules into smaller molecules, such as glucose. It is used to measure the efficiency of enzyme-catalyzed reactions and the digestibility of cellulose. The enzymatic digestibility of a compound is affected by factors such as the compound's crystallinity index, the type of enzyme used, and the temperature and pH of the reaction. In this study the enzymatic digestibility of the alkali pretreated SB was improved by increasing pretreatment conditions. Typically the pretreatment settings are selected by considering various factors such as feedstock characteristics, pretreatment chemical cost, energy consumption and recovery efficiency (Wu et al. 2011). α-cellulose obtained by 8% alkali treated of SB was used for hydrolysis reaction and the reaction was carried out at an optimum condition set by Sujan et al. (2018). The most effective enzymatic hydrolysis was taken place with Trichoderma virideat 48 h and the theoretical yield of sugars i.e. glucose and xylose were obtained 89.59% and 61.23%, respectively. Through fermentation process generated sugars were used to check ethanol production. Yeast Saccharomyces cerevisiae presented worthy performance to convert C6 sugar into ethanol when it was incubated at 30°C for 24 h and 16.81±0.32% ethanol yield was detected. Conclusion α-cellulose yield was optimized by varying alkali concentration, temperature and cooking time. The most vital influencing factor for α-cellulose yield was alkali concentration and after that temperature performed a little bit. The optimum α-cellulose yield (63.58±0.05%) was obtained at 8% alkali concentration, 100oC and 90 minutes. In hydrolysis step of SB, Trichoderma viridaewas used to convert 8% alkali treated SB into sugars and it was attained 89.59% glucose and 61.23% xylose which gave higher yield in compare with our previous study such as ball milled and mesh size varied SB sample (Sujan et al. 2018). At fermentation step Saccha- romyces cerevisiae was used to convert hydrolysate of 8% alkali treated SB sample and ethanol yield was obtained 16.81±0.32% at 24 h. Compositional analysis, imaging and crystallinity are three methods were performed. SEM images can give different clues about SB including morphology, surface disruption and creation of highly accessible surface area iii) CrI (18.68% to 39.19%) and crystal size (3.28 nm to 4.06 nm) of SB are increased with different alkali treatment (0%-8%). In this case it was observed that crystallinity, crystal size, accessible surface area, porosity, particle size, lignin and hemicellulose content and enzyme adsorption/de- sorption were acted as the most impressive factors for digest- ibility of sugarcane bagasse. Acknowledgement Authors would like to acknowledge Pulp and Paper Division, BCSIR for their cooperation. Authors also like to thanks for the sincere assistance of Md. Abdul Hamid, Junior Techni- cian, IFRD, BCSIR. References Alvira P, Tomás-Pejó E, Ballesteros M and Negro MJ (2010), Pretreatment technologies for an efficient bioethanol production process based on enzymatic hydrolysis: a review, Bioresource technology 101(13): 4851-4861. DOI: org/10.1016/j.biortech.2009.11.093 Agbor VB, Cicek N, Sparling R, Berlin A and Levin, DB (2011), Biomass pretreatment: fundamentals toward application, Biotechnology advances 29(6): 675-685. DOI: org/10.1016/j.biotechadv.2011.05.005 Amiri H and Karimi K (2015), Improvement of acetone, butanol, and ethanol production from woody biomass using organosolv pretreatment, Bioprocess and biosys- tems engineering 38(10): 1959-1972. ASTM (2007), Standard test members for instrumental deter- mination of C, H, N in petroleum products and lubricants. ASTM D5291-02, ASTM international, West Conshohocken, PA. Canilha L, Chandel AK, Suzane dos Santos Milessi T, Antunes FAF, Luiz da Costa Freitas W, das Graças Almeida Felipe M and da Silva SS (2012), Bioconver- sion of sugarcane biomass into ethanol: an overview about composition, pretreatment methods, detoxifica- tion of hydrolysates, enzymatic saccharification and ethanol fermentation. Journal of Biomedicine and Biotechnology, DOI: org/10.1155/2012/989572 Den W, Sharma VK, Lee M, Nadadur G and Varma RS (2018), Lignocellulosic biomass transformations via greener oxidative pretreatment processes: access to energy and value-added chemicals, Frontiers in chem- istry 6: 141. DOI: org/10.3389/fchem.2018.00141 Ahmed FM, Rahman SR and Gomes DJ (2012), Saccharifi- cation of sugarcane bagasse by enzymatic treatment for bioethanol production, Malays J Microbiol 8(2), 97-103. Hirani AH, Javed N, Asif M, Basu SK and Kumar A (2018), A review on first-and second-generation biofuel productions. In Biofuels: greenhouse gas mitigation and global warming, Springer, New Delhi, pp 141-154. DOI: 10.1007/978-81-322-3763-1_8 Islam MZ, Asad MA, Hossain MT, Paul SC and Sujan SMA (2019), Bioethanol production from banana pseudostem by using separate and cocultures of cellu- lase enzyme with Saccharomyces cerevisiae, Journal of Environmental Science and Technology 12(4): 157-163. Jahan MS, Saeed A, Ni Y and He Z (2009), Pre-extraction and its impact on the alkaline pulping of bagasse, Journal of Biobased Materials and Bioenergy 3(4): 380-385. DOI: org/10.1166/jbmb.2009.1053 Janker-Obermeier I, Sieber V, Faulstich Mand Schieder D (2012), Solubilization of hemicellulose and lignin from wheat straw through microwave-assisted alkali treatment, Industrial Crops and Products 39: 198-203. DOI: org/10.1016/j.indcrop.2012.02.022 Karimi K and Taherzadeh MJ (2016), A critical review of analytical methods in pretreatment of lignocelluloses: composition, imaging, and crystallinity, Bioresource technology 200: 1008-1018. DOI: org/10.1016/j. biortech.2015.11.022 Maryana R, Ma’rifatun D, Wheni AI, Satriyo KW and Rizal WA (2014), Alkaline pretreatment on sugarcane bagasse for bioethanol production, Energy Procedia, 47: 250-254. DOI: org/10.1016/j.egypro.2014.01.221 McIntosh Sand Vancov T (2010), Enhanced enzyme sacchar- ification of Sorghum bicolor straw using dilute alkali pretreatment, Bioresource technology 101(17): 6718-6727. DOI: org/10.1016/j.biortech.2010.03.116 Meyer KH and Misch L (1937), Positions des atomes dans le nouveau modele spatial de la cellulose. Helvetica Chimica Acta 20(1): 232-244. DOI: org/10. 1002/hlca.19370200134 Park S, Baker JO, Himmel ME, Parilla PA and Johnson DK (2010), Cellulose crystallinity index: measurement techniques and their impact on interpreting cellulase performance, Biotechnology for biofuels 3(1): 10. Sujan SMA, Bari ML and Fakhruddin AN (2018), Effects of physical pretreatment (crushing and ball milling) on sugarcane bagasse for bioethanol production, Bangla- desh journal of botany 147(2): 257-64. Wilkie JS (1961), Carl Nägeli and the fine structure of living matter, Nature 190: 1145-1150. Wu L, Arakane M, Ike M, Wada M, Takai T, Gau Mand Tokuyasu K (2011), Low temperature alkali pretreat- ment for improving enzymatic digestibility of sweet sorghum bagasse for ethanol production, Bioresource Technology 102(7): 4793-4799. Zhang W, Okubayashi S and Bechtold T (2005), Fibrillation tendency of cellulosic fibers-Part 4. Effects of alkali pretreatment of various cellulosic fibers, Carbohydrate polymers. 61(4): 427-433. Zhao X, van der Heide E, Zhang Tand Liu D (2010), Delig- nification of sugarcane bagasse with alkali and peracetic acid and characterization of the pulp, BioRe- sources 5(3): 1565-1580. Zhao X, Zhang Land Liu D (2012), Biomass recalcitrance. Part II: Fundamentals of different pre‐treatments to increase the enzymatic digestibility of lignocellulose, Biofuels, Bioproducts and Biorefining 6(5): 561-579. DOI: https://doi.org/10.3329/bjsir.v58i2.66990 Received: 01 February 2023 Revised: 04 April 2023 Accepted: 11 April 2023 Optimization of alkali concentration in the pretreatment of sugarcane bagasse 58(2) 202390 composed of vasculer bundles surrounded by non-fibrous parenchymatic cell (Jahan et al. 2009). Usually it is available as agricultural residue and byproduct from sugar mill indus- tries. In the year of 2015-16, there was 2,95,162 metric ton sugarcane bagasse was obtained from sugar mills of Bangla- desh (http://www.bsfic.gov.bd). To enrich cellulose content in biomass for enzymatic saccarification, current research has focused on various pretreatment processes. Currently, scientists have gone through a lot of studies on different pretreatment techniques to remove the compact and rigid composition by open up the cellulosic structure (Alvira et al. 2010). Pretreatment technologies are commonly abandoned to reduce the structural barriers and boost cellulose availabili- ty based on not only chemicals such as acid, alkali, oxidant, etc. but also several treatment settings. Alkali treatment is one of the most widespread and economical methods used for surface modification of lignocellulosic biomass. In the field of biorefinary, alkali pretreatment is intensively employed to develop the cellulosic materials both mechan- ically as well as chemically and such properties include tensile strength, dyeability, stability of dimension and reactivity (Wu et al. 2011). In alkali pretreatment biomass is treated under moderate reaction conditions ensuring inexpensiveness, inflated recycling possibilities of water and chemical agents (McIntosh and Vancov, 2010). Usually alkali treatment is the most effective pretreatment for agricultural residues, herbaceous crops and hardwood containing low lignin content in the compariosn of softwood containing high lignin content (Agbor et al. 2011; Canilha et al. 2012). This research aims is to investigate the alkaline pretreatment repercussions based on the chemical structure, surface morphology, structure and enzymatic digestion of sugarcane bagasse for sugars and ethanol production. Materials and methods Sugarcane bagasse (SB) was taken from street juice vendor. Warm tap water was used for washing purpose to eliminate the residual free sugars once again. This washing and pressing step was then repeated for three times. After that drying was done in an oven at 65oC for 16 h. Crushing of dried bagasse was done successively using a locally made crusher and sieved (Retsch, D-42759, HAAN, Germany) to have the particle size of 20-40 mesh on average. Before usage, the 20-40 mesh bagasse was kept at room temperature in an airtight plastic container (Sujan et al. 2018). Moisture and Ash content Moisture content of raw materials was measured using the ASTM D 4442-07 procedure. In a glass crucible, 2 grams of pretreatment sample containing 20-40 mesh were dried in oven at 105±2°C. The moisture content was expressed in percent wet basis and weight measurements were taken every 3 h. A muffle furnace was used for burning the Oven-dried samples at 575±25°C for ash content determination follow- ing the ASTM Standard E 1755-01. Volatile matter The volatile matter analysis was carried out in accordance with ASTM Standard D 271-48. Four grams of raw materials were heated in a furnace at 950±20°C for seven minutes. The weight loss, excluding the weight of moisture pushed off at 105°C, is then used to calculate volatile matter. Fixed carbon The difference between 100 and the total of volatile matter, moisture, and ash content was used to compute the fixed carbon percentage. Chemical analysis The technical association of the pulp and paper industry (TAPPI) method detects α-cellulose (T 203 cm-99), pentosan (T 223 cm-01), klason lignin (T211 om-83), and acid soluble lignin (T UM 250) on a dry basis. Ultimate analysis Ultimate analysis of samples was done by following the procedure ASTM Standard D 5291-02. Organic elemental analyzer (Flash 2000, Thermo Scientific, USA) was used with a specific condition (Reactor temp. 900°C, He: 250 kPa, O2: 250 kPa, TCD). Low concentration alkali pretreatment Different alkali concentration (0%, 2%, 4%, 6% and 8%), time (60, 90 and 120 min) and temperature (80, 100 and 120oC) were applied on raw material to check alkali effect on SB shown in Table 3. A portion of the ground bagasse 20-40 mesh (10g) was taken into plastic zipper bag (temperature 80oC, figure 2a and 2b) and stainless-steel reactor (tempera- ture 100oC and 120oC, figure 2c and 2d). Different alkali concentration was applied during the pretreatment process (alkali to SB ratio ranging between 1:12). The mixture was heated at a particular temperature such as in a water bath (80oC) and oil bath (100oC and 120oC) for a desired length of time. In the course of pretreatment process the sample was manually mixed 2-3 times to attain proper alkali treatment. The treated SB was placed into a polyester bag to remove excess alkali water by pressing it. After that it was vigorously washed with tap water (repeated for five times) to remove the remaining alkali. Finally the SB was dried in an oven at 65oC for 72 h and stored in a close container at room temperature for further experiment. Regression model Regression model has been developed for prediction of α-cel- lulose, pentosan and lignin in SB. The general form of the model is: y=α+α1x1 + α2x2+........................+αnxn + ε.......................(1) where α is the constant term. αi are the coefficient of variables xi. ε is the random error term which is minimized with Simple Least Squares Regression (SLSR). Regression coefficients of the independent variables namely alkali concentration, temperature and time are estimated by SLSR method for developing regression model to predict α-cellulose, pentosan and lignin. Efficiencies of these models are expressed by coefficient of multiple determination (R2) and Adjusted R2. 2.9 Separate hydrolysis and fermentation (SHF) The Hydrolysis experiment took place in 100 ml conical flux 10ml enzyme solution with 200 mg (2% dry wt.) in citrate buffer (0.05 M, pH 5.0) at 50°C for 48 h. In this case Trichoderma viride was used for hydrolysis. Hydrolysate was then heated for 15 min in a boiling water bath and centrifugation was done to remove solid particles. The supernatant was used for analysis of released sugars as described by Jamal et al. (2011). During fermentation process according to Firoz et al. (2012), 100 ml media was prepared and 0.5 g of commercial yeast Saccharomyces cerevisiae was used as inoculum which showed good performance to converts sugar into bioethanol. This inoculated media mixture was poured in a suitable glass- ware and was kept in a shaking incubator for 48 h. 10 ml of this medium was then added into the flask and it was properly covered with aluminum foil. Then it was placed in the incubator at 30oC for 24 h, 48 h, 72 h and 96 h for fermenta- tion of sugars to bio-ethanol according to Sujan et al. (2018). Samples from hydrolysis and fermentation were performed by HPLC. High performance liquid chromatography (HPLC) In characterize part, concentration of Sugars and ethanol were determined by HPLC (Ultimate 3000, Thermo Scien- tific, USA) method using Hyper Rez XP carbohydrate H+ 8 µm column (100×7.7 mm) equipped with a Refractive Index (Shodex RI-101) detector. The mobile phase was degassed with deionized water with a flow rate of 0.7 ml/min and column temperature was maintained at 70°C. It is possible to measure the total sugar concentration in the hydrolysis liquid fraction by comparing its peak area detected by HPLC with peak area of 1% standard sugar which consists of two sugars namely glucose and xylose (Sujan et al. 2018). The same column which is specialized for fermentation broth analysis is used for ethanol detec- tion. The kinetic parameters of ethanol fermentation were determined as follows (Islam et al. 2019): Crystallinity measurement X-ray diffraction (XRD) was used to determine the crystalline structure of the SB samples using a diffrac- tometer (GBC XRD) and filtered copper K radiation (λ = 0.1542 nm) by a monochromator at 35.50 kV voltage and 28 mA current, with a speed of about 2o/min and scan- ning in the range of 10 - 80ºC. The crystallinity index (CrI) was obtained from the ratio between the intensity of the 002 peak (I002, 2θ = 22.5) and the minimum dip (Iam, 2 θ = 18.5) according to the following equation (Roberta et al., 2012): CrI (%) = [(I002 - Iam)/I002] ×100 ...................................... (2) where I002 is the highest peak intensity of plane 002 and Iam is related to the amorphous structure. In present study, the average crystallite sizes were deter- mined from the Scherrer equation by using the diffraction pattern obtained from the 002 (hkl) lattice planes of cellulose samples D(hkl) = [(Kλ / B(hkl) cos2θ] ............................................... (3) Where D(hkl) (Crystallite size), K (Scherrer constant, 0.84), λ (X-ray wavelength, 0.154nm), B(hkl)(Full width half maxi- mum of the measured hkl reflection), and 2θ (Corresponding Bragg angle). Scanning electron microscopy (SEM) analysis In this research, SEM (ZEISS EVO 18 SEM) was used to detect the change of pretreated bagasse fibers. SEM images were taken of different pretreated bagasse samples with acceleration voltage of 2.0 KV. Results and discussion Proximate analysis Proximate analysis of SB sample (20-40 mesh) are presented in Table I. Primarily this analysis usually evaluate the fuel characteristics of raw materials. According to Sun et al., 2009, higher moisture and ashcontent in samples lessen the heating value. Ultimate analysis Ultimate analysis denotes the elemental configuration of SB such as carbon, hydrogen, oxygen, nitrogen and sulfur which are shown in Table II. This examination helps to measure the percentage of carbon and hydrogen content in biomass that is responsible to determine the amount of air is required for complete combustion, composition of combustion gases and heat is generated by it (Poddar et al. 2014). Chemical properties Raw SB contained 34.66±12% α-cellulose, 22.43±08% pentosan and 19.57±06% klason lignin in which 1.75±04% acid soluble lignin (dry basis) was detected by technical association of the pulp and paper industry (TAPPI) method. The chemical composition of SB was determined by acid hydrolysis and it was calculatedby HPLC method as 45.35% glucose and 30.64% xylose (Sujan et al., 2018). α-cellulose yield Bagasse is mainly composed of cellulose, hemicellulose and lignin. Besides these there are some extractives such as ash, wax, gum, pectin etc. During alkali pretreatment usually most of the extractives are removed with the increasing of alkali concentration. Pretreatment of SB with different alkali concentration based on raw material (0%, 2%, 4%, 6% and 8%), time (60, 90 and 120 min) and temperature (80, 100 and 120oC) are shown in Table III. Consequences of each independent experiment varying with alkali concentration, temperature and time on the α-cellulose yield were analyzed using MATLAB software. Apparently, it was observed that α-cellulose yield was considerably increased with changes of alkali concentration ranging from 0% to 8% (34.66% to 63.58%). But no noticeable variation was observed in temperature-time alteration during pretreatment of SB. Based on cellulose percentage obtained in treated bagasse, the optimum conditions for pretreatment reaction were selected as alkali concentration 8%, time 90 min and temperature 100oC. Although cellulose content in treated SB was slowly increased with the increase of alkali concentration 12% attime 90 min and temperature 100oC but as a consequence of huge chemical consumption, recovery problem, chance of losses cellulose and hemicellulose, alkali concentrations for pretreatment above 8% was not considered as ideal concentration. Effect of alkali concentration (AC), cooking time and temperature (temp) charge on α-cellulose yield as well as their statistical significance on the basis of F-test number are presented in regression equations (4). As shown in equations, cooking time at the maximum temperature had no significant effect on α-cellulose yield followed by alkali concentration charge. Effect of temperature on α-cellulose yield was less in employed cooking conditions. For α-cellulose yield: α-cellulose yield = 82.05-0.151×temp-1.56×time-0.69×AC (R2=0.89, adjusted R2=0.87) ............................................. (4) For Pentosan: Pentosan = 13.58+0.086 × tem p + 0.006 × time+0.445 × AC (R2=0.67, adjusted R2=0.63) ............................................. (5) For Lignin: Lignin = 27.662-0.041×temp-0.001×time-01.242×AC (R2=0.93, adjusted R2=0.92) ............................................. (6) For predicting α-cellulose, percentage of pentosan and lignin, the most influential factor was alkali concentration and then cooking temperature for α-cellulose yield, which exhibited an almost linear dependence on both operational variables. The coefficient of determinations is good for α-cel- lulose yield and lignin percentage which hovers around 90 percent, although the figure is moderate more than 60 percent for pentosan. All these three models are significant (p<0.05) at 5% level of significance. In order to perceive the impact of alkali concentration and temperature on these three parameters, three-dimensional (3D) response surface plots were created by plotting the response (α-cellulose yield) pentosan and lignin on the Z-axis versus the most influential one independent variable alkali concentration and temperature as shown in Fig. 3 (a), (b) and (c). XRD analysis In 19th century the cellulose crystalline structure has been discovered and later it was verified by X-ray crystallography (Meyer and Misch, 1937; Wilkie, 1961). The crystallinity index (CrI) of non-woody biomass, such as grasses and agricultural residues, varies depending on the type of biomass. Generally, non-woody biomass has a lower crystallinity index than woody biomass, typically ranging from 20-40%. The crystallinity index of non-woody biomass affects its digestibility and energy production potential.Recently, researchers are being paid more atten- tion on cellulose index because of its potential use in bioen- ergy production. Since then several different models of cellulose index have been proposed. The most popular two-phase cellulose model describes cellulose chains as containing both crystalline (ordered) and amorphous (less ordered) region (Park et al., 2010). Alkali treatment of bagasse has been found to increase the crystallinity index of the bagasse. Alkali treatment breaks down hemicellulose, making the cellulose molecules more ordered and crystalline, resulting in a higher crystallinity index. Alkali treatment also increases the digestibility of the bagasse, making it easier to break down and therefore more suitable for bioenergy production. X-ray diffraction spectra of the SB and alkali treated SB were presented in Figure 4. It was observed that the intensity of 101 and 002 peaks were gradually increased. The relative amount of crystalline cellu- lose (CrI) in the total solid were calculated based on the equation (2) and obtained 18.68%, 24.80%, 27.23%, 32.69%, 36.45% and 39.19% for raw SB, 0%, 2%, 4%, 6% and 8% alkali treated SB samples respectively. The CrI of alkali treated SB samples (not cellulose crystallinity) were intense- ly influenced by the composition of the samples. In case of lignocellulosic biomass examples, cellulose CrI measured the relative amount of crystalline cellulose in the total solid. Therefore, amorphous part of lignin and hemicellulose in biomass specimens were partially removed with the delig- nification process as a result the proportion of α-cellulose was increased and hence CrI would be increased gradually. This interpretation could be proved by the fact that alkali-treated SB had higher CrI than raw SB (Zhao et al. 2010) and the portion of cellulose in the treated SB was also increased gradually. According to the equation 3, the average sizes of crystallite obtained were 3.28, 3.51, 3.63, 3.78, 3.90 and 4.06 nm for raw SB, 0%, 2%, 4%, 6% and 8% alkali treated SB samples respectively. The experimental data revealed that during delignification, the size of crystallite was increased. SEM analysis SEM is one of the most commanding tools widely used to inves- tigate the lignocellulosic biomass surface (Amiri and Karimi, 2015). SEM is usually employed for surface characterization, morphology and inspection of microstructure. In respect of biomass example through SEM images we can compare the untreated and pretreated models which may lead to different insight into the biomass (Karimi and Taherzadeh, 2016). SEM provides two-dimensional images of raw SB and alkali treated SB, which were taken and compared to the outcome of NaOH treatment. All testers were coated with carbon tape and magnification of 500x was used. Figure 4 shows the wall of raw SB (Figure 6a) was intact where the alkali treated SB (Figures 6c-6f), the cell wall was ruptured or splitted and hence packing of the fibers were partially loosened (Firoz et al. 2012). Sugar and ethanol yield Enzymatic digestibility is the ability of enzymes to break down molecules into smaller molecules, such as glucose. It is used to measure the efficiency of enzyme-catalyzed reactions and the digestibility of cellulose. The enzymatic digestibility of a compound is affected by factors such as the compound's crystallinity index, the type of enzyme used, and the temperature and pH of the reaction. In this study the enzymatic digestibility of the alkali pretreated SB was improved by increasing pretreatment conditions. Typically the pretreatment settings are selected by considering various factors such as feedstock characteristics, pretreatment chemical cost, energy consumption and recovery efficiency (Wu et al. 2011). α-cellulose obtained by 8% alkali treated of SB was used for hydrolysis reaction and the reaction was carried out at an optimum condition set by Sujan et al. (2018). The most effective enzymatic hydrolysis was taken place with Trichoderma virideat 48 h and the theoretical yield of sugars i.e. glucose and xylose were obtained 89.59% and 61.23%, respectively. Through fermentation process generated sugars were used to check ethanol production. Yeast Saccharomyces cerevisiae presented worthy performance to convert C6 sugar into ethanol when it was incubated at 30°C for 24 h and 16.81±0.32% ethanol yield was detected. Conclusion α-cellulose yield was optimized by varying alkali concentration, temperature and cooking time. The most vital influencing factor for α-cellulose yield was alkali concentration and after that temperature performed a little bit. The optimum α-cellulose yield (63.58±0.05%) was obtained at 8% alkali concentration, 100oC and 90 minutes. In hydrolysis step of SB, Trichoderma viridaewas used to convert 8% alkali treated SB into sugars and it was attained 89.59% glucose and 61.23% xylose which gave higher yield in compare with our previous study such as ball milled and mesh size varied SB sample (Sujan et al. 2018). At fermentation step Saccha- romyces cerevisiae was used to convert hydrolysate of 8% alkali treated SB sample and ethanol yield was obtained 16.81±0.32% at 24 h. Compositional analysis, imaging and crystallinity are three methods were performed. SEM images can give different clues about SB including morphology, surface disruption and creation of highly accessible surface area iii) CrI (18.68% to 39.19%) and crystal size (3.28 nm to 4.06 nm) of SB are increased with different alkali treatment (0%-8%). In this case it was observed that crystallinity, crystal size, accessible surface area, porosity, particle size, lignin and hemicellulose content and enzyme adsorption/de- sorption were acted as the most impressive factors for digest- ibility of sugarcane bagasse. Acknowledgement Authors would like to acknowledge Pulp and Paper Division, BCSIR for their cooperation. Authors also like to thanks for the sincere assistance of Md. Abdul Hamid, Junior Techni- cian, IFRD, BCSIR. References Alvira P, Tomás-Pejó E, Ballesteros M and Negro MJ (2010), Pretreatment technologies for an efficient bioethanol production process based on enzymatic hydrolysis: a review, Bioresource technology 101(13): 4851-4861. DOI: org/10.1016/j.biortech.2009.11.093 Agbor VB, Cicek N, Sparling R, Berlin A and Levin, DB (2011), Biomass pretreatment: fundamentals toward application, Biotechnology advances 29(6): 675-685. DOI: org/10.1016/j.biotechadv.2011.05.005 Amiri H and Karimi K (2015), Improvement of acetone, butanol, and ethanol production from woody biomass using organosolv pretreatment, Bioprocess and biosys- tems engineering 38(10): 1959-1972. ASTM (2007), Standard test members for instrumental deter- mination of C, H, N in petroleum products and lubricants. ASTM D5291-02, ASTM international, West Conshohocken, PA. Canilha L, Chandel AK, Suzane dos Santos Milessi T, Antunes FAF, Luiz da Costa Freitas W, das Graças Almeida Felipe M and da Silva SS (2012), Bioconver- sion of sugarcane biomass into ethanol: an overview about composition, pretreatment methods, detoxifica- tion of hydrolysates, enzymatic saccharification and ethanol fermentation. Journal of Biomedicine and Biotechnology, DOI: org/10.1155/2012/989572 Den W, Sharma VK, Lee M, Nadadur G and Varma RS (2018), Lignocellulosic biomass transformations via greener oxidative pretreatment processes: access to energy and value-added chemicals, Frontiers in chem- istry 6: 141. DOI: org/10.3389/fchem.2018.00141 Ahmed FM, Rahman SR and Gomes DJ (2012), Saccharifi- cation of sugarcane bagasse by enzymatic treatment for bioethanol production, Malays J Microbiol 8(2), 97-103. Hirani AH, Javed N, Asif M, Basu SK and Kumar A (2018), A review on first-and second-generation biofuel productions. In Biofuels: greenhouse gas mitigation and global warming, Springer, New Delhi, pp 141-154. DOI: 10.1007/978-81-322-3763-1_8 Islam MZ, Asad MA, Hossain MT, Paul SC and Sujan SMA (2019), Bioethanol production from banana pseudostem by using separate and cocultures of cellu- lase enzyme with Saccharomyces cerevisiae, Journal of Environmental Science and Technology 12(4): 157-163. Jahan MS, Saeed A, Ni Y and He Z (2009), Pre-extraction and its impact on the alkaline pulping of bagasse, Journal of Biobased Materials and Bioenergy 3(4): 380-385. DOI: org/10.1166/jbmb.2009.1053 Janker-Obermeier I, Sieber V, Faulstich Mand Schieder D (2012), Solubilization of hemicellulose and lignin from wheat straw through microwave-assisted alkali treatment, Industrial Crops and Products 39: 198-203. DOI: org/10.1016/j.indcrop.2012.02.022 Karimi K and Taherzadeh MJ (2016), A critical review of analytical methods in pretreatment of lignocelluloses: composition, imaging, and crystallinity, Bioresource technology 200: 1008-1018. DOI: org/10.1016/j. biortech.2015.11.022 Maryana R, Ma’rifatun D, Wheni AI, Satriyo KW and Rizal WA (2014), Alkaline pretreatment on sugarcane bagasse for bioethanol production, Energy Procedia, 47: 250-254. DOI: org/10.1016/j.egypro.2014.01.221 McIntosh Sand Vancov T (2010), Enhanced enzyme sacchar- ification of Sorghum bicolor straw using dilute alkali pretreatment, Bioresource technology 101(17): 6718-6727. DOI: org/10.1016/j.biortech.2010.03.116 Meyer KH and Misch L (1937), Positions des atomes dans le nouveau modele spatial de la cellulose. Helvetica Chimica Acta 20(1): 232-244. DOI: org/10. 1002/hlca.19370200134 Park S, Baker JO, Himmel ME, Parilla PA and Johnson DK (2010), Cellulose crystallinity index: measurement techniques and their impact on interpreting cellulase performance, Biotechnology for biofuels 3(1): 10. Sujan SMA, Bari ML and Fakhruddin AN (2018), Effects of physical pretreatment (crushing and ball milling) on sugarcane bagasse for bioethanol production, Bangla- desh journal of botany 147(2): 257-64. Wilkie JS (1961), Carl Nägeli and the fine structure of living matter, Nature 190: 1145-1150. Wu L, Arakane M, Ike M, Wada M, Takai T, Gau Mand Tokuyasu K (2011), Low temperature alkali pretreat- ment for improving enzymatic digestibility of sweet sorghum bagasse for ethanol production, Bioresource Technology 102(7): 4793-4799. Zhang W, Okubayashi S and Bechtold T (2005), Fibrillation tendency of cellulosic fibers-Part 4. Effects of alkali pretreatment of various cellulosic fibers, Carbohydrate polymers. 61(4): 427-433. Zhao X, van der Heide E, Zhang Tand Liu D (2010), Delig- nification of sugarcane bagasse with alkali and peracetic acid and characterization of the pulp, BioRe- sources 5(3): 1565-1580. Zhao X, Zhang Land Liu D (2012), Biomass recalcitrance. Part II: Fundamentals of different pre‐treatments to increase the enzymatic digestibility of lignocellulose, Biofuels, Bioproducts and Biorefining 6(5): 561-579. Fig. 1. 20-40 mesh size SB Sujan, Hossain, Uddin and Fakhruddin 91 composed of vasculer bundles surrounded by non-fibrous parenchymatic cell (Jahan et al. 2009). Usually it is available as agricultural residue and byproduct from sugar mill indus- tries. In the year of 2015-16, there was 2,95,162 metric ton sugarcane bagasse was obtained from sugar mills of Bangla- desh (http://www.bsfic.gov.bd). To enrich cellulose content in biomass for enzymatic saccarification, current research has focused on various pretreatment processes. Currently, scientists have gone through a lot of studies on different pretreatment techniques to remove the compact and rigid composition by open up the cellulosic structure (Alvira et al. 2010). Pretreatment technologies are commonly abandoned to reduce the structural barriers and boost cellulose availabili- ty based on not only chemicals such as acid, alkali, oxidant, etc. but also several treatment settings. Alkali treatment is one of the most widespread and economical methods used for surface modification of lignocellulosic biomass. In the field of biorefinary, alkali pretreatment is intensively employed to develop the cellulosic materials both mechan- ically as well as chemically and such properties include tensile strength, dyeability, stability of dimension and reactivity (Wu et al. 2011). In alkali pretreatment biomass is treated under moderate reaction conditions ensuring inexpensiveness, inflated recycling possibilities of water and chemical agents (McIntosh and Vancov, 2010). Usually alkali treatment is the most effective pretreatment for agricultural residues, herbaceous crops and hardwood containing low lignin content in the compariosn of softwood containing high lignin content (Agbor et al. 2011; Canilha et al. 2012). This research aims is to investigate the alkaline pretreatment repercussions based on the chemical structure, surface morphology, structure and enzymatic digestion of sugarcane bagasse for sugars and ethanol production. Materials and methods Sugarcane bagasse (SB) was taken from street juice vendor. Warm tap water was used for washing purpose to eliminate the residual free sugars once again. This washing and pressing step was then repeated for three times. After that drying was done in an oven at 65oC for 16 h. Crushing of dried bagasse was done successively using a locally made crusher and sieved (Retsch, D-42759, HAAN, Germany) to have the particle size of 20-40 mesh on average. Before usage, the 20-40 mesh bagasse was kept at room temperature in an airtight plastic container (Sujan et al. 2018). Moisture and Ash content Moisture content of raw materials was measured using the ASTM D 4442-07 procedure. In a glass crucible, 2 grams of pretreatment sample containing 20-40 mesh were dried in oven at 105±2°C. The moisture content was expressed in percent wet basis and weight measurements were taken every 3 h. A muffle furnace was used for burning the Oven-dried samples at 575±25°C for ash content determination follow- ing the ASTM Standard E 1755-01. Volatile matter The volatile matter analysis was carried out in accordance with ASTM Standard D 271-48. Four grams of raw materials were heated in a furnace at 950±20°C for seven minutes. The weight loss, excluding the weight of moisture pushed off at 105°C, is then used to calculate volatile matter. Fixed carbon The difference between 100 and the total of volatile matter, moisture, and ash content was used to compute the fixed carbon percentage. Chemical analysis The technical association of the pulp and paper industry (TAPPI) method detects α-cellulose (T 203 cm-99), pentosan (T 223 cm-01), klason lignin (T211 om-83), and acid soluble lignin (T UM 250) on a dry basis. Ultimate analysis Ultimate analysis of samples was done by following the procedure ASTM Standard D 5291-02. Organic elemental analyzer (Flash 2000, Thermo Scientific, USA) was used with a specific condition (Reactor temp. 900°C, He: 250 kPa, O2: 250 kPa, TCD). Low concentration alkali pretreatment Different alkali concentration (0%, 2%, 4%, 6% and 8%), time (60, 90 and 120 min) and temperature (80, 100 and 120oC) were applied on raw material to check alkali effect on SB shown in Table 3. A portion of the ground bagasse 20-40 mesh (10g) was taken into plastic zipper bag (temperature 80oC, figure 2a and 2b) and stainless-steel reactor (tempera- ture 100oC and 120oC, figure 2c and 2d). Different alkali concentration was applied during the pretreatment process (alkali to SB ratio ranging between 1:12). The mixture was heated at a particular temperature such as in a water bath (80oC) and oil bath (100oC and 120oC) for a desired length of time. In the course of pretreatment process the sample was manually mixed 2-3 times to attain proper alkali treatment. The treated SB was placed into a polyester bag to remove excess alkali water by pressing it. After that it was vigorously washed with tap water (repeated for five times) to remove the remaining alkali. Finally the SB was dried in an oven at 65oC for 72 h and stored in a close container at room temperature for further experiment. Regression model Regression model has been developed for prediction of α-cel- lulose, pentosan and lignin in SB. The general form of the model is: y=α+α1x1 + α2x2+........................+αnxn + ε.......................(1) where α is the constant term. αi are the coefficient of variables xi. ε is the random error term which is minimized with Simple Least Squares Regression (SLSR). Regression coefficients of the independent variables namely alkali concentration, temperature and time are estimated by SLSR method for developing regression model to predict α-cellulose, pentosan and lignin. Efficiencies of these models are expressed by coefficient of multiple determination (R2) and Adjusted R2. 2.9 Separate hydrolysis and fermentation (SHF) The Hydrolysis experiment took place in 100 ml conical flux 10ml enzyme solution with 200 mg (2% dry wt.) in citrate buffer (0.05 M, pH 5.0) at 50°C for 48 h. In this case Trichoderma viride was used for hydrolysis. Hydrolysate was then heated for 15 min in a boiling water bath and centrifugation was done to remove solid particles. The supernatant was used for analysis of released sugars as described by Jamal et al. (2011). During fermentation process according to Firoz et al. (2012), 100 ml media was prepared and 0.5 g of commercial yeast Saccharomyces cerevisiae was used as inoculum which showed good performance to converts sugar into bioethanol. This inoculated media mixture was poured in a suitable glass- ware and was kept in a shaking incubator for 48 h. 10 ml of this medium was then added into the flask and it was properly covered with aluminum foil. Then it was placed in the incubator at 30oC for 24 h, 48 h, 72 h and 96 h for fermenta- tion of sugars to bio-ethanol according to Sujan et al. (2018). Samples from hydrolysis and fermentation were performed by HPLC. High performance liquid chromatography (HPLC) In characterize part, concentration of Sugars and ethanol were determined by HPLC (Ultimate 3000, Thermo Scien- tific, USA) method using Hyper Rez XP carbohydrate H+ 8 µm column (100×7.7 mm) equipped with a Refractive Index (Shodex RI-101) detector. The mobile phase was degassed with deionized water with a flow rate of 0.7 ml/min and column temperature was maintained at 70°C. It is possible to measure the total sugar concentration in the hydrolysis liquid fraction by comparing its peak area detected by HPLC with peak area of 1% standard sugar which consists of two sugars namely glucose and xylose (Sujan et al. 2018). The same column which is specialized for fermentation broth analysis is used for ethanol detec- tion. The kinetic parameters of ethanol fermentation were determined as follows (Islam et al. 2019): Crystallinity measurement X-ray diffraction (XRD) was used to determine the crystalline structure of the SB samples using a diffrac- tometer (GBC XRD) and filtered copper K radiation (λ = 0.1542 nm) by a monochromator at 35.50 kV voltage and 28 mA current, with a speed of about 2o/min and scan- ning in the range of 10 - 80ºC. The crystallinity index (CrI) was obtained from the ratio between the intensity of the 002 peak (I002, 2θ = 22.5) and the minimum dip (Iam, 2 θ = 18.5) according to the following equation (Roberta et al., 2012): CrI (%) = [(I002 - Iam)/I002] ×100 ...................................... (2) where I002 is the highest peak intensity of plane 002 and Iam is related to the amorphous structure. In present study, the average crystallite sizes were deter- mined from the Scherrer equation by using the diffraction pattern obtained from the 002 (hkl) lattice planes of cellulose samples D(hkl) = [(Kλ / B(hkl) cos2θ] ............................................... (3) Where D(hkl) (Crystallite size), K (Scherrer constant, 0.84), λ (X-ray wavelength, 0.154nm), B(hkl)(Full width half maxi- mum of the measured hkl reflection), and 2θ (Corresponding Bragg angle). Scanning electron microscopy (SEM) analysis In this research, SEM (ZEISS EVO 18 SEM) was used to detect the change of pretreated bagasse fibers. SEM images were taken of different pretreated bagasse samples with acceleration voltage of 2.0 KV. Results and discussion Proximate analysis Proximate analysis of SB sample (20-40 mesh) are presented in Table I. Primarily this analysis usually evaluate the fuel characteristics of raw materials. According to Sun et al., 2009, higher moisture and ashcontent in samples lessen the heating value. Ultimate analysis Ultimate analysis denotes the elemental configuration of SB such as carbon, hydrogen, oxygen, nitrogen and sulfur which are shown in Table II. This examination helps to measure the percentage of carbon and hydrogen content in biomass that is responsible to determine the amount of air is required for complete combustion, composition of combustion gases and heat is generated by it (Poddar et al. 2014). Chemical properties Raw SB contained 34.66±12% α-cellulose, 22.43±08% pentosan and 19.57±06% klason lignin in which 1.75±04% acid soluble lignin (dry basis) was detected by technical association of the pulp and paper industry (TAPPI) method. The chemical composition of SB was determined by acid hydrolysis and it was calculatedby HPLC method as 45.35% glucose and 30.64% xylose (Sujan et al., 2018). α-cellulose yield Bagasse is mainly composed of cellulose, hemicellulose and lignin. Besides these there are some extractives such as ash, wax, gum, pectin etc. During alkali pretreatment usually most of the extractives are removed with the increasing of alkali concentration. Pretreatment of SB with different alkali concentration based on raw material (0%, 2%, 4%, 6% and 8%), time (60, 90 and 120 min) and temperature (80, 100 and 120oC) are shown in Table III. Consequences of each independent experiment varying with alkali concentration, temperature and time on the α-cellulose yield were analyzed using MATLAB software. Apparently, it was observed that α-cellulose yield was considerably increased with changes of alkali concentration ranging from 0% to 8% (34.66% to 63.58%). But no noticeable variation was observed in temperature-time alteration during pretreatment of SB. Based on cellulose percentage obtained in treated bagasse, the optimum conditions for pretreatment reaction were selected as alkali concentration 8%, time 90 min and temperature 100oC. Although cellulose content in treated SB was slowly increased with the increase of alkali concentration 12% attime 90 min and temperature 100oC but as a consequence of huge chemical consumption, recovery problem, chance of losses cellulose and hemicellulose, alkali concentrations for pretreatment above 8% was not considered as ideal concentration. Effect of alkali concentration (AC), cooking time and temperature (temp) charge on α-cellulose yield as well as their statistical significance on the basis of F-test number are presented in regression equations (4). As shown in equations, cooking time at the maximum temperature had no significant effect on α-cellulose yield followed by alkali concentration charge. Effect of temperature on α-cellulose yield was less in employed cooking conditions. For α-cellulose yield: α-cellulose yield = 82.05-0.151×temp-1.56×time-0.69×AC (R2=0.89, adjusted R2=0.87) ............................................. (4) For Pentosan: Pentosan = 13.58+0.086 × tem p + 0.006 × time+0.445 × AC (R2=0.67, adjusted R2=0.63) ............................................. (5) For Lignin: Lignin = 27.662-0.041×temp-0.001×time-01.242×AC (R2=0.93, adjusted R2=0.92) ............................................. (6) For predicting α-cellulose, percentage of pentosan and lignin, the most influential factor was alkali concentration and then cooking temperature for α-cellulose yield, which exhibited an almost linear dependence on both operational variables. The coefficient of determinations is good for α-cel- lulose yield and lignin percentage which hovers around 90 percent, although the figure is moderate more than 60 percent for pentosan. All these three models are significant (p<0.05) at 5% level of significance. In order to perceive the impact of alkali concentration and temperature on these three parameters, three-dimensional (3D) response surface plots were created by plotting the response (α-cellulose yield) pentosan and lignin on the Z-axis versus the most influential one independent variable alkali concentration and temperature as shown in Fig. 3 (a), (b) and (c). XRD analysis In 19th century the cellulose crystalline structure has been discovered and later it was verified by X-ray crystallography (Meyer and Misch, 1937; Wilkie, 1961). The crystallinity index (CrI) of non-woody biomass, such as grasses and agricultural residues, varies depending on the type of biomass. Generally, non-woody biomass has a lower crystallinity index than woody biomass, typically ranging from 20-40%. The crystallinity index of non-woody biomass affects its digestibility and energy production potential.Recently, researchers are being paid more atten- tion on cellulose index because of its potential use in bioen- ergy production. Since then several different models of cellulose index have been proposed. The most popular two-phase cellulose model describes cellulose chains as containing both crystalline (ordered) and amorphous (less ordered) region (Park et al., 2010). Alkali treatment of bagasse has been found to increase the crystallinity index of the bagasse. Alkali treatment breaks down hemicellulose, making the cellulose molecules more ordered and crystalline, resulting in a higher crystallinity index. Alkali treatment also increases the digestibility of the bagasse, making it easier to break down and therefore more suitable for bioenergy production. X-ray diffraction spectra of the SB and alkali treated SB were presented in Figure 4. It was observed that the intensity of 101 and 002 peaks were gradually increased. The relative amount of crystalline cellu- lose (CrI) in the total solid were calculated based on the equation (2) and obtained 18.68%, 24.80%, 27.23%, 32.69%, 36.45% and 39.19% for raw SB, 0%, 2%, 4%, 6% and 8% alkali treated SB samples respectively. The CrI of alkali treated SB samples (not cellulose crystallinity) were intense- ly influenced by the composition of the samples. In case of lignocellulosic biomass examples, cellulose CrI measured the relative amount of crystalline cellulose in the total solid. Therefore, amorphous part of lignin and hemicellulose in biomass specimens were partially removed with the delig- nification process as a result the proportion of α-cellulose was increased and hence CrI would be increased gradually. This interpretation could be proved by the fact that alkali-treated SB had higher CrI than raw SB (Zhao et al. 2010) and the portion of cellulose in the treated SB was also increased gradually. According to the equation 3, the average sizes of crystallite obtained were 3.28, 3.51, 3.63, 3.78, 3.90 and 4.06 nm for raw SB, 0%, 2%, 4%, 6% and 8% alkali treated SB samples respectively. The experimental data revealed that during delignification, the size of crystallite was increased. SEM analysis SEM is one of the most commanding tools widely used to inves- tigate the lignocellulosic biomass surface (Amiri and Karimi, 2015). SEM is usually employed for surface characterization, morphology and inspection of microstructure. In respect of biomass example through SEM images we can compare the untreated and pretreated models which may lead to different insight into the biomass (Karimi and Taherzadeh, 2016). SEM provides two-dimensional images of raw SB and alkali treated SB, which were taken and compared to the outcome of NaOH treatment. All testers were coated with carbon tape and magnification of 500x was used. Figure 4 shows the wall of raw SB (Figure 6a) was intact where the alkali treated SB (Figures 6c-6f), the cell wall was ruptured or splitted and hence packing of the fibers were partially loosened (Firoz et al. 2012). Sugar and ethanol yield Enzymatic digestibility is the ability of enzymes to break down molecules into smaller molecules, such as glucose. It is used to measure the efficiency of enzyme-catalyzed reactions and the digestibility of cellulose. The enzymatic digestibility of a compound is affected by factors such as the compound's crystallinity index, the type of enzyme used, and the temperature and pH of the reaction. In this study the enzymatic digestibility of the alkali pretreated SB was improved by increasing pretreatment conditions. Typically the pretreatment settings are selected by considering various factors such as feedstock characteristics, pretreatment chemical cost, energy consumption and recovery efficiency (Wu et al. 2011). α-cellulose obtained by 8% alkali treated of SB was used for hydrolysis reaction and the reaction was carried out at an optimum condition set by Sujan et al. (2018). The most effective enzymatic hydrolysis was taken place with Trichoderma virideat 48 h and the theoretical yield of sugars i.e. glucose and xylose were obtained 89.59% and 61.23%, respectively. Through fermentation process generated sugars were used to check ethanol production. Yeast Saccharomyces cerevisiae presented worthy performance to convert C6 sugar into ethanol when it was incubated at 30°C for 24 h and 16.81±0.32% ethanol yield was detected. Conclusion α-cellulose yield was optimized by varying alkali concentration, temperature and cooking time. The most vital influencing factor for α-cellulose yield was alkali concentration and after that temperature performed a little bit. The optimum α-cellulose yield (63.58±0.05%) was obtained at 8% alkali concentration, 100oC and 90 minutes. In hydrolysis step of SB, Trichoderma viridaewas used to convert 8% alkali treated SB into sugars and it was attained 89.59% glucose and 61.23% xylose which gave higher yield in compare with our previous study such as ball milled and mesh size varied SB sample (Sujan et al. 2018). At fermentation step Saccha- romyces cerevisiae was used to convert hydrolysate of 8% alkali treated SB sample and ethanol yield was obtained 16.81±0.32% at 24 h. Compositional analysis, imaging and crystallinity are three methods were performed. SEM images can give different clues about SB including morphology, surface disruption and creation of highly accessible surface area iii) CrI (18.68% to 39.19%) and crystal size (3.28 nm to 4.06 nm) of SB are increased with different alkali treatment (0%-8%). In this case it was observed that crystallinity, crystal size, accessible surface area, porosity, particle size, lignin and hemicellulose content and enzyme adsorption/de- sorption were acted as the most impressive factors for digest- ibility of sugarcane bagasse. Acknowledgement Authors would like to acknowledge Pulp and Paper Division, BCSIR for their cooperation. Authors also like to thanks for the sincere assistance of Md. Abdul Hamid, Junior Techni- cian, IFRD, BCSIR. References Alvira P, Tomás-Pejó E, Ballesteros M and Negro MJ (2010), Pretreatment technologies for an efficient bioethanol production process based on enzymatic hydrolysis: a review, Bioresource technology 101(13): 4851-4861. DOI: org/10.1016/j.biortech.2009.11.093 Agbor VB, Cicek N, Sparling R, Berlin A and Levin, DB (2011), Biomass pretreatment: fundamentals toward application, Biotechnology advances 29(6): 675-685. DOI: org/10.1016/j.biotechadv.2011.05.005 Amiri H and Karimi K (2015), Improvement of acetone, butanol, and ethanol production from woody biomass using organosolv pretreatment, Bioprocess and biosys- tems engineering 38(10): 1959-1972. ASTM (2007), Standard test members for instrumental deter- mination of C, H, N in petroleum products and lubricants. ASTM D5291-02, ASTM international, West Conshohocken, PA. Canilha L, Chandel AK, Suzane dos Santos Milessi T, Antunes FAF, Luiz da Costa Freitas W, das Graças Almeida Felipe M and da Silva SS (2012), Bioconver- sion of sugarcane biomass into ethanol: an overview about composition, pretreatment methods, detoxifica- tion of hydrolysates, enzymatic saccharification and ethanol fermentation. Journal of Biomedicine and Biotechnology, DOI: org/10.1155/2012/989572 Den W, Sharma VK, Lee M, Nadadur G and Varma RS (2018), Lignocellulosic biomass transformations via greener oxidative pretreatment processes: access to energy and value-added chemicals, Frontiers in chem- istry 6: 141. DOI: org/10.3389/fchem.2018.00141 Ahmed FM, Rahman SR and Gomes DJ (2012), Saccharifi- cation of sugarcane bagasse by enzymatic treatment for bioethanol production, Malays J Microbiol 8(2), 97-103. Hirani AH, Javed N, Asif M, Basu SK and Kumar A (2018), A review on first-and second-generation biofuel productions. In Biofuels: greenhouse gas mitigation and global warming, Springer, New Delhi, pp 141-154. DOI: 10.1007/978-81-322-3763-1_8 Islam MZ, Asad MA, Hossain MT, Paul SC and Sujan SMA (2019), Bioethanol production from banana pseudostem by using separate and cocultures of cellu- lase enzyme with Saccharomyces cerevisiae, Journal of Environmental Science and Technology 12(4): 157-163. Jahan MS, Saeed A, Ni Y and He Z (2009), Pre-extraction and its impact on the alkaline pulping of bagasse, Journal of Biobased Materials and Bioenergy 3(4): 380-385. DOI: org/10.1166/jbmb.2009.1053 Janker-Obermeier I, Sieber V, Faulstich Mand Schieder D (2012), Solubilization of hemicellulose and lignin from wheat straw through microwave-assisted alkali treatment, Industrial Crops and Products 39: 198-203. DOI: org/10.1016/j.indcrop.2012.02.022 Karimi K and Taherzadeh MJ (2016), A critical review of analytical methods in pretreatment of lignocelluloses: composition, imaging, and crystallinity, Bioresource technology 200: 1008-1018. DOI: org/10.1016/j. biortech.2015.11.022 Maryana R, Ma’rifatun D, Wheni AI, Satriyo KW and Rizal WA (2014), Alkaline pretreatment on sugarcane bagasse for bioethanol production, Energy Procedia, 47: 250-254. DOI: org/10.1016/j.egypro.2014.01.221 McIntosh Sand Vancov T (2010), Enhanced enzyme sacchar- ification of Sorghum bicolor straw using dilute alkali pretreatment, Bioresource technology 101(17): 6718-6727. DOI: org/10.1016/j.biortech.2010.03.116 Meyer KH and Misch L (1937), Positions des atomes dans le nouveau modele spatial de la cellulose. Helvetica Chimica Acta 20(1): 232-244. DOI: org/10. 1002/hlca.19370200134 Park S, Baker JO, Himmel ME, Parilla PA and Johnson DK (2010), Cellulose crystallinity index: measurement techniques and their impact on interpreting cellulase performance, Biotechnology for biofuels 3(1): 10. Sujan SMA, Bari ML and Fakhruddin AN (2018), Effects of physical pretreatment (crushing and ball milling) on sugarcane bagasse for bioethanol production, Bangla- desh journal of botany 147(2): 257-64. Wilkie JS (1961), Carl Nägeli and the fine structure of living matter, Nature 190: 1145-1150. Wu L, Arakane M, Ike M, Wada M, Takai T, Gau Mand Tokuyasu K (2011), Low temperature alkali pretreat- ment for improving enzymatic digestibility of sweet sorghum bagasse for ethanol production, Bioresource Technology 102(7): 4793-4799. Zhang W, Okubayashi S and Bechtold T (2005), Fibrillation tendency of cellulosic fibers-Part 4. Effects of alkali pretreatment of various cellulosic fibers, Carbohydrate polymers. 61(4): 427-433. Zhao X, van der Heide E, Zhang Tand Liu D (2010), Delig- nification of sugarcane bagasse with alkali and peracetic acid and characterization of the pulp, BioRe- sources 5(3): 1565-1580. Zhao X, Zhang Land Liu D (2012), Biomass recalcitrance. Part II: Fundamentals of different pre‐treatments to increase the enzymatic digestibility of lignocellulose, Biofuels, Bioproducts and Biorefining 6(5): 561-579. Fig. 2. a) 20-40 mesh size SB and alkali were taking into plastic zipper bag; b) samples were pretreated into water bath at a temperature below 100oC; c) 20-40 mesh size SB and alkali were taking into stainless-steel reactor; d) sample pretreated into oil bath at a temperature above 100oC 2a 2b 2c 2d Optimization of alkali concentration in the pretreatment of sugarcane bagasse 58(2) 202392 composed of vasculer bundles surrounded by non-fibrous parenchymatic cell (Jahan et al. 2009). Usually it is available as agricultural residue and byproduct from sugar mill indus- tries. In the year of 2015-16, there was 2,95,162 metric ton sugarcane bagasse was obtained from sugar mills of Bangla- desh (http://www.bsfic.gov.bd). To enrich cellulose content in biomass for enzymatic saccarification, current research has focused on various pretreatment processes. Currently, scientists have gone through a lot of studies on different pretreatment techniques to remove the compact and rigid composition by open up the cellulosic structure (Alvira et al. 2010). Pretreatment technologies are commonly abandoned to reduce the structural barriers and boost cellulose availabili- ty based on not only chemicals such as acid, alkali, oxidant, etc. but also several treatment settings. Alkali treatment is one of the most widespread and economical methods used for surface modification of lignocellulosic biomass. In the field of biorefinary, alkali pretreatment is intensively employed to develop the cellulosic materials both mechan- ically as well as chemically and such properties include tensile strength, dyeability, stability of dimension and reactivity (Wu et al. 2011). In alkali pretreatment biomass is treated under moderate reaction conditions ensuring inexpensiveness, inflated recycling possibilities of water and chemical agents (McIntosh and Vancov, 2010). Usually alkali treatment is the most effective pretreatment for agricultural residues, herbaceous crops and hardwood containing low lignin content in the compariosn of softwood containing high lignin content (Agbor et al. 2011; Canilha et al. 2012). This research aims is to investigate the alkaline pretreatment repercussions based on the chemical structure, surface morphology, structure and enzymatic digestion of sugarcane bagasse for sugars and ethanol production. Materials and methods Sugarcane bagasse (SB) was taken from street juice vendor. Warm tap water was used for washing purpose to eliminate the residual free sugars once again. This washing and pressing step was then repeated for three times. After that drying was done in an oven at 65oC for 16 h. Crushing of dried bagasse was done successively using a locally made crusher and sieved (Retsch, D-42759, HAAN, Germany) to have the particle size of 20-40 mesh on average. Before usage, the 20-40 mesh bagasse was kept at room temperature in an airtight plastic container (Sujan et al. 2018). Moisture and Ash content Moisture content of raw materials was measured using the ASTM D 4442-07 procedure. In a glass crucible, 2 grams of pretreatment sample containing 20-40 mesh were dried in oven at 105±2°C. The moisture content was expressed in percent wet basis and weight measurements were taken every 3 h. A muffle furnace was used for burning the Oven-dried samples at 575±25°C for ash content determination follow- ing the ASTM Standard E 1755-01. Volatile matter The volatile matter analysis was carried out in accordance with ASTM Standard D 271-48. Four grams of raw materials were heated in a furnace at 950±20°C for seven minutes. The weight loss, excluding the weight of moisture pushed off at 105°C, is then used to calculate volatile matter. Fixed carbon The difference between 100 and the total of volatile matter, moisture, and ash content was used to compute the fixed carbon percentage. Chemical analysis The technical association of the pulp and paper industry (TAPPI) method detects α-cellulose (T 203 cm-99), pentosan (T 223 cm-01), klason lignin (T211 om-83), and acid soluble lignin (T UM 250) on a dry basis. Ultimate analysis Ultimate analysis of samples was done by following the procedure ASTM Standard D 5291-02. Organic elemental analyzer (Flash 2000, Thermo Scientific, USA) was used with a specific condition (Reactor temp. 900°C, He: 250 kPa, O2: 250 kPa, TCD). Low concentration alkali pretreatment Different alkali concentration (0%, 2%, 4%, 6% and 8%), time (60, 90 and 120 min) and temperature (80, 100 and 120oC) were applied on raw material to check alkali effect on SB shown in Table 3. A portion of the ground bagasse 20-40 mesh (10g) was taken into plastic zipper bag (temperature 80oC, figure 2a and 2b) and stainless-steel reactor (tempera- ture 100oC and 120oC, figure 2c and 2d). Different alkali concentration was applied during the pretreatment process (alkali to SB ratio ranging between 1:12). The mixture was heated at a particular temperature such as in a water bath (80oC) and oil bath (100oC and 120oC) for a desired length of time. In the course of pretreatment process the sample was manually mixed 2-3 times to attain proper alkali treatment. The treated SB was placed into a polyester bag to remove excess alkali water by pressing it. After that it was vigorously washed with tap water (repeated for five times) to remove the remaining alkali. Finally the SB was dried in an oven at 65oC for 72 h and stored in a close container at room temperature for further experiment. Regression model Regression model has been developed for prediction of α-cel- lulose, pentosan and lignin in SB. The general form of the model is: y=α+α1x1 + α2x2+........................+αnxn + ε.......................(1) where α is the constant term. αi are the coefficient of variables xi. ε is the random error term which is minimized with Simple Least Squares Regression (SLSR). Regression coefficients of the independent variables namely alkali concentration, temperature and time are estimated by SLSR method for developing regression model to predict α-cellulose, pentosan and lignin. Efficiencies of these models are expressed by coefficient of multiple determination (R2) and Adjusted R2. 2.9 Separate hydrolysis and fermentation (SHF) The Hydrolysis experiment took place in 100 ml conical flux 10ml enzyme solution with 200 mg (2% dry wt.) in citrate buffer (0.05 M, pH 5.0) at 50°C for 48 h. In this case Trichoderma viride was used for hydrolysis. Hydrolysate was then heated for 15 min in a boiling water bath and centrifugation was done to remove solid particles. The supernatant was used for analysis of released sugars as described by Jamal et al. (2011). During fermentation process according to Firoz et al. (2012), 100 ml media was prepared and 0.5 g of commercial yeast Saccharomyces cerevisiae was used as inoculum which showed good performance to converts sugar into bioethanol. This inoculated media mixture was poured in a suitable glass- ware and was kept in a shaking incubator for 48 h. 10 ml of this medium was then added into the flask and it was properly covered with aluminum foil. Then it was placed in the incubator at 30oC for 24 h, 48 h, 72 h and 96 h for fermenta- tion of sugars to bio-ethanol according to Sujan et al. (2018). Samples from hydrolysis and fermentation were performed by HPLC. High performance liquid chromatography (HPLC) In characterize part, concentration of Sugars and ethanol were determined by HPLC (Ultimate 3000, Thermo Scien- tific, USA) method using Hyper Rez XP carbohydrate H+ 8 µm column (100×7.7 mm) equipped with a Refractive Index (Shodex RI-101) detector. The mobile phase was degassed with deionized water with a flow rate of 0.7 ml/min and column temperature was maintained at 70°C. It is possible to measure the total sugar concentration in the hydrolysis liquid fraction by comparing its peak area detected by HPLC with peak area of 1% standard sugar which consists of two sugars namely glucose and xylose (Sujan et al. 2018). The same column which is specialized for fermentation broth analysis is used for ethanol detec- tion. The kinetic parameters of ethanol fermentation were determined as follows (Islam et al. 2019): Crystallinity measurement X-ray diffraction (XRD) was used to determine the crystalline structure of the SB samples using a diffrac- tometer (GBC XRD) and filtered copper K radiation (λ = 0.1542 nm) by a monochromator at 35.50 kV voltage and 28 mA current, with a speed of about 2o/min and scan- ning in the range of 10 - 80ºC. The crystallinity index (CrI) was obtained from the ratio between the intensity of the 002 peak (I002, 2θ = 22.5) and the minimum dip (Iam, 2 θ = 18.5) according to the following equation (Roberta et al., 2012): CrI (%) = [(I002 - Iam)/I002] ×100 ...................................... (2) where I002 is the highest peak intensity of plane 002 and Iam is related to the amorphous structure. In present study, the average crystallite sizes were deter- mined from the Scherrer equation by using the diffraction pattern obtained from the 002 (hkl) lattice planes of cellulose samples D(hkl) = [(Kλ / B(hkl) cos2θ] ............................................... (3) Where D(hkl) (Crystallite size), K (Scherrer constant, 0.84), λ (X-ray wavelength, 0.154nm), B(hkl)(Full width half maxi- mum of the measured hkl reflection), and 2θ (Corresponding Bragg angle). Scanning electron microscopy (SEM) analysis In this research, SEM (ZEISS EVO 18 SEM) was used to detect the change of pretreated bagasse fibers. SEM images were taken of different pretreated bagasse samples with acceleration voltage of 2.0 KV. Results and discussion Proximate analysis Proximate analysis of SB sample (20-40 mesh) are presented in Table I. Primarily this analysis usually evaluate the fuel characteristics of raw materials. According to Sun et al., 2009, higher moisture and ashcontent in samples lessen the heating value. Ultimate analysis Ultimate analysis denotes the elemental configuration of SB such as carbon, hydrogen, oxygen, nitrogen and sulfur which are shown in Table II. This examination helps to measure the percentage of carbon and hydrogen content in biomass that is responsible to determine the amount of air is required for complete combustion, composition of combustion gases and heat is generated by it (Poddar et al. 2014). Chemical properties Raw SB contained 34.66±12% α-cellulose, 22.43±08% pentosan and 19.57±06% klason lignin in which 1.75±04% acid soluble lignin (dry basis) was detected by technical association of the pulp and paper industry (TAPPI) method. The chemical composition of SB was determined by acid hydrolysis and it was calculatedby HPLC method as 45.35% glucose and 30.64% xylose (Sujan et al., 2018). α-cellulose yield Bagasse is mainly composed of cellulose, hemicellulose and lignin. Besides these there are some extractives such as ash, wax, gum, pectin etc. During alkali pretreatment usually most of the extractives are removed with the increasing of alkali concentration. Pretreatment of SB with different alkali concentration based on raw material (0%, 2%, 4%, 6% and 8%), time (60, 90 and 120 min) and temperature (80, 100 and 120oC) are shown in Table III. Consequences of each independent experiment varying with alkali concentration, temperature and time on the α-cellulose yield were analyzed using MATLAB software. Apparently, it was observed that α-cellulose yield was considerably increased with changes of alkali concentration ranging from 0% to 8% (34.66% to 63.58%). But no noticeable variation was observed in temperature-time alteration during pretreatment of SB. Based on cellulose percentage obtained in treated bagasse, the optimum conditions for pretreatment reaction were selected as alkali concentration 8%, time 90 min and temperature 100oC. Although cellulose content in treated SB was slowly increased with the increase of alkali concentration 12% attime 90 min and temperature 100oC but as a consequence of huge chemical consumption, recovery problem, chance of losses cellulose and hemicellulose, alkali concentrations for pretreatment above 8% was not considered as ideal concentration. Effect of alkali concentration (AC), cooking time and temperature (temp) charge on α-cellulose yield as well as their statistical significance on the basis of F-test number are presented in regression equations (4). As shown in equations, cooking time at the maximum temperature had no significant effect on α-cellulose yield followed by alkali concentration charge. Effect of temperature on α-cellulose yield was less in employed cooking conditions. For α-cellulose yield: α-cellulose yield = 82.05-0.151×temp-1.56×time-0.69×AC (R2=0.89, adjusted R2=0.87) ............................................. (4) For Pentosan: Pentosan = 13.58+0.086 × tem p + 0.006 × time+0.445 × AC (R2=0.67, adjusted R2=0.63) ............................................. (5) For Lignin: Lignin = 27.662-0.041×temp-0.001×time-01.242×AC (R2=0.93, adjusted R2=0.92) ............................................. (6) For predicting α-cellulose, percentage of pentosan and lignin, the most influential factor was alkali concentration and then cooking temperature for α-cellulose yield, which exhibited an almost linear dependence on both operational variables. The coefficient of determinations is good for α-cel- lulose yield and lignin percentage which hovers around 90 percent, although the figure is moderate more than 60 percent for pentosan. All these three models are significant (p<0.05) at 5% level of significance. In order to perceive the impact of alkali concentration and temperature on these three parameters, three-dimensional (3D) response surface plots were created by plotting the response (α-cellulose yield) pentosan and lignin on the Z-axis versus the most influential one independent variable alkali concentration and temperature as shown in Fig. 3 (a), (b) and (c). XRD analysis In 19th century the cellulose crystalline structure has been discovered and later it was verified by X-ray crystallography (Meyer and Misch, 1937; Wilkie, 1961). The crystallinity index (CrI) of non-woody biomass, such as grasses and agricultural residues, varies depending on the type of biomass. Generally, non-woody biomass has a lower crystallinity index than woody biomass, typically ranging from 20-40%. The crystallinity index of non-woody biomass affects its digestibility and energy production potential.Recently, researchers are being paid more atten- tion on cellulose index because of its potential use in bioen- ergy production. Since then several different models of cellulose index have been proposed. The most popular two-phase cellulose model describes cellulose chains as containing both crystalline (ordered) and amorphous (less ordered) region (Park et al., 2010). Alkali treatment of bagasse has been found to increase the crystallinity index of the bagasse. Alkali treatment breaks down hemicellulose, making the cellulose molecules more ordered and crystalline, resulting in a higher crystallinity index. Alkali treatment also increases the digestibility of the bagasse, making it easier to break down and therefore more suitable for bioenergy production. X-ray diffraction spectra of the SB and alkali treated SB were presented in Figure 4. It was observed that the intensity of 101 and 002 peaks were gradually increased. The relative amount of crystalline cellu- lose (CrI) in the total solid were calculated based on the equation (2) and obtained 18.68%, 24.80%, 27.23%, 32.69%, 36.45% and 39.19% for raw SB, 0%, 2%, 4%, 6% and 8% alkali treated SB samples respectively. The CrI of alkali treated SB samples (not cellulose crystallinity) were intense- ly influenced by the composition of the samples. In case of lignocellulosic biomass examples, cellulose CrI measured the relative amount of crystalline cellulose in the total solid. Therefore, amorphous part of lignin and hemicellulose in biomass specimens were partially removed with the delig- nification process as a result the proportion of α-cellulose was increased and hence CrI would be increased gradually. This interpretation could be proved by the fact that alkali-treated SB had higher CrI than raw SB (Zhao et al. 2010) and the portion of cellulose in the treated SB was also increased gradually. According to the equation 3, the average sizes of crystallite obtained were 3.28, 3.51, 3.63, 3.78, 3.90 and 4.06 nm for raw SB, 0%, 2%, 4%, 6% and 8% alkali treated SB samples respectively. The experimental data revealed that during delignification, the size of crystallite was increased. SEM analysis SEM is one of the most commanding tools widely used to inves- tigate the lignocellulosic biomass surface (Amiri and Karimi, 2015). SEM is usually employed for surface characterization, morphology and inspection of microstructure. In respect of biomass example through SEM images we can compare the untreated and pretreated models which may lead to different insight into the biomass (Karimi and Taherzadeh, 2016). SEM provides two-dimensional images of raw SB and alkali treated SB, which were taken and compared to the outcome of NaOH treatment. All testers were coated with carbon tape and magnification of 500x was used. Figure 4 shows the wall of raw SB (Figure 6a) was intact where the alkali treated SB (Figures 6c-6f), the cell wall was ruptured or splitted and hence packing of the fibers were partially loosened (Firoz et al. 2012). Sugar and ethanol yield Enzymatic digestibility is the ability of enzymes to break down molecules into smaller molecules, such as glucose. It is used to measure the efficiency of enzyme-catalyzed reactions and the digestibility of cellulose. The enzymatic digestibility of a compound is affected by factors such as the compound's crystallinity index, the type of enzyme used, and the temperature and pH of the reaction. In this study the enzymatic digestibility of the alkali pretreated SB was improved by increasing pretreatment conditions. Typically the pretreatment settings are selected by considering various factors such as feedstock characteristics, pretreatment chemical cost, energy consumption and recovery efficiency (Wu et al. 2011). α-cellulose obtained by 8% alkali treated of SB was used for hydrolysis reaction and the reaction was carried out at an optimum condition set by Sujan et al. (2018). The most effective enzymatic hydrolysis was taken place with Trichoderma virideat 48 h and the theoretical yield of sugars i.e. glucose and xylose were obtained 89.59% and 61.23%, respectively. Through fermentation process generated sugars were used to check ethanol production. Yeast Saccharomyces cerevisiae presented worthy performance to convert C6 sugar into ethanol when it was incubated at 30°C for 24 h and 16.81±0.32% ethanol yield was detected. Conclusion α-cellulose yield was optimized by varying alkali concentration, temperature and cooking time. The most vital influencing factor for α-cellulose yield was alkali concentration and after that temperature performed a little bit. The optimum α-cellulose yield (63.58±0.05%) was obtained at 8% alkali concentration, 100oC and 90 minutes. In hydrolysis step of SB, Trichoderma viridaewas used to convert 8% alkali treated SB into sugars and it was attained 89.59% glucose and 61.23% xylose which gave higher yield in compare with our previous study such as ball milled and mesh size varied SB sample (Sujan et al. 2018). At fermentation step Saccha- romyces cerevisiae was used to convert hydrolysate of 8% alkali treated SB sample and ethanol yield was obtained 16.81±0.32% at 24 h. Compositional analysis, imaging and crystallinity are three methods were performed. SEM images can give different clues about SB including morphology, surface disruption and creation of highly accessible surface area iii) CrI (18.68% to 39.19%) and crystal size (3.28 nm to 4.06 nm) of SB are increased with different alkali treatment (0%-8%). In this case it was observed that crystallinity, crystal size, accessible surface area, porosity, particle size, lignin and hemicellulose content and enzyme adsorption/de- sorption were acted as the most impressive factors for digest- ibility of sugarcane bagasse. Acknowledgement Authors would like to acknowledge Pulp and Paper Division, BCSIR for their cooperation. Authors also like to thanks for the sincere assistance of Md. Abdul Hamid, Junior Techni- cian, IFRD, BCSIR. References Alvira P, Tomás-Pejó E, Ballesteros M and Negro MJ (2010), Pretreatment technologies for an efficient bioethanol production process based on enzymatic hydrolysis: a review, Bioresource technology 101(13): 4851-4861. DOI: org/10.1016/j.biortech.2009.11.093 Agbor VB, Cicek N, Sparling R, Berlin A and Levin, DB (2011), Biomass pretreatment: fundamentals toward application, Biotechnology advances 29(6): 675-685. DOI: org/10.1016/j.biotechadv.2011.05.005 Amiri H and Karimi K (2015), Improvement of acetone, butanol, and ethanol production from woody biomass using organosolv pretreatment, Bioprocess and biosys- tems engineering 38(10): 1959-1972. ASTM (2007), Standard test members for instrumental deter- mination of C, H, N in petroleum products and lubricants. ASTM D5291-02, ASTM international, West Conshohocken, PA. Canilha L, Chandel AK, Suzane dos Santos Milessi T, Antunes FAF, Luiz da Costa Freitas W, das Graças Almeida Felipe M and da Silva SS (2012), Bioconver- sion of sugarcane biomass into ethanol: an overview about composition, pretreatment methods, detoxifica- tion of hydrolysates, enzymatic saccharification and ethanol fermentation. Journal of Biomedicine and Biotechnology, DOI: org/10.1155/2012/989572 Den W, Sharma VK, Lee M, Nadadur G and Varma RS (2018), Lignocellulosic biomass transformations via greener oxidative pretreatment processes: access to energy and value-added chemicals, Frontiers in chem- istry 6: 141. DOI: org/10.3389/fchem.2018.00141 Ahmed FM, Rahman SR and Gomes DJ (2012), Saccharifi- cation of sugarcane bagasse by enzymatic treatment for bioethanol production, Malays J Microbiol 8(2), 97-103. Hirani AH, Javed N, Asif M, Basu SK and Kumar A (2018), A review on first-and second-generation biofuel productions. In Biofuels: greenhouse gas mitigation and global warming, Springer, New Delhi, pp 141-154. DOI: 10.1007/978-81-322-3763-1_8 Islam MZ, Asad MA, Hossain MT, Paul SC and Sujan SMA (2019), Bioethanol production from banana pseudostem by using separate and cocultures of cellu- lase enzyme with Saccharomyces cerevisiae, Journal of Environmental Science and Technology 12(4): 157-163. Jahan MS, Saeed A, Ni Y and He Z (2009), Pre-extraction and its impact on the alkaline pulping of bagasse, Journal of Biobased Materials and Bioenergy 3(4): 380-385. DOI: org/10.1166/jbmb.2009.1053 Janker-Obermeier I, Sieber V, Faulstich Mand Schieder D (2012), Solubilization of hemicellulose and lignin from wheat straw through microwave-assisted alkali treatment, Industrial Crops and Products 39: 198-203. DOI: org/10.1016/j.indcrop.2012.02.022 Karimi K and Taherzadeh MJ (2016), A critical review of analytical methods in pretreatment of lignocelluloses: composition, imaging, and crystallinity, Bioresource technology 200: 1008-1018. DOI: org/10.1016/j. biortech.2015.11.022 Maryana R, Ma’rifatun D, Wheni AI, Satriyo KW and Rizal WA (2014), Alkaline pretreatment on sugarcane bagasse for bioethanol production, Energy Procedia, 47: 250-254. DOI: org/10.1016/j.egypro.2014.01.221 McIntosh Sand Vancov T (2010), Enhanced enzyme sacchar- ification of Sorghum bicolor straw using dilute alkali pretreatment, Bioresource technology 101(17): 6718-6727. DOI: org/10.1016/j.biortech.2010.03.116 Meyer KH and Misch L (1937), Positions des atomes dans le nouveau modele spatial de la cellulose. Helvetica Chimica Acta 20(1): 232-244. DOI: org/10. 1002/hlca.19370200134 Park S, Baker JO, Himmel ME, Parilla PA and Johnson DK (2010), Cellulose crystallinity index: measurement techniques and their impact on interpreting cellulase performance, Biotechnology for biofuels 3(1): 10. Sujan SMA, Bari ML and Fakhruddin AN (2018), Effects of physical pretreatment (crushing and ball milling) on sugarcane bagasse for bioethanol production, Bangla- desh journal of botany 147(2): 257-64. Wilkie JS (1961), Carl Nägeli and the fine structure of living matter, Nature 190: 1145-1150. Wu L, Arakane M, Ike M, Wada M, Takai T, Gau Mand Tokuyasu K (2011), Low temperature alkali pretreat- ment for improving enzymatic digestibility of sweet sorghum bagasse for ethanol production, Bioresource Technology 102(7): 4793-4799. Zhang W, Okubayashi S and Bechtold T (2005), Fibrillation tendency of cellulosic fibers-Part 4. Effects of alkali pretreatment of various cellulosic fibers, Carbohydrate polymers. 61(4): 427-433. Zhao X, van der Heide E, Zhang Tand Liu D (2010), Delig- nification of sugarcane bagasse with alkali and peracetic acid and characterization of the pulp, BioRe- sources 5(3): 1565-1580. Zhao X, Zhang Land Liu D (2012), Biomass recalcitrance. Part II: Fundamentals of different pre‐treatments to increase the enzymatic digestibility of lignocellulose, Biofuels, Bioproducts and Biorefining 6(5): 561-579. Ethanol concentration (Ec)% = Volume of reaction mixture (L) Ethanol produced (g) Table I. Proximate analysis of sugarcane bagasse sample Name of sample Moisture (%) Ash (%) Volatile matter (%) Fixed carbon (%) SB 9.44±0.14 1.75±0.06 4.99±0.11 83.82±0.17 Sujan, Hossain, Uddin and Fakhruddin 93 composed of vasculer bundles surrounded by non-fibrous parenchymatic cell (Jahan et al. 2009). Usually it is available as agricultural residue and byproduct from sugar mill indus- tries. In the year of 2015-16, there was 2,95,162 metric ton sugarcane bagasse was obtained from sugar mills of Bangla- desh (http://www.bsfic.gov.bd). To enrich cellulose content in biomass for enzymatic saccarification, current research has focused on various pretreatment processes. Currently, scientists have gone through a lot of studies on different pretreatment techniques to remove the compact and rigid composition by open up the cellulosic structure (Alvira et al. 2010). Pretreatment technologies are commonly abandoned to reduce the structural barriers and boost cellulose availabili- ty based on not only chemicals such as acid, alkali, oxidant, etc. but also several treatment settings. Alkali treatment is one of the most widespread and economical methods used for surface modification of lignocellulosic biomass. In the field of biorefinary, alkali pretreatment is intensively employed to develop the cellulosic materials both mechan- ically as well as chemically and such properties include tensile strength, dyeability, stability of dimension and reactivity (Wu et al. 2011). In alkali pretreatment biomass is treated under moderate reaction conditions ensuring inexpensiveness, inflated recycling possibilities of water and chemical agents (McIntosh and Vancov, 2010). Usually alkali treatment is the most effective pretreatment for agricultural residues, herbaceous crops and hardwood containing low lignin content in the compariosn of softwood containing high lignin content (Agbor et al. 2011; Canilha et al. 2012). This research aims is to investigate the alkaline pretreatment repercussions based on the chemical structure, surface morphology, structure and enzymatic digestion of sugarcane bagasse for sugars and ethanol production. Materials and methods Sugarcane bagasse (SB) was taken from street juice vendor. Warm tap water was used for washing purpose to eliminate the residual free sugars once again. This washing and pressing step was then repeated for three times. After that drying was done in an oven at 65oC for 16 h. Crushing of dried bagasse was done successively using a locally made crusher and sieved (Retsch, D-42759, HAAN, Germany) to have the particle size of 20-40 mesh on average. Before usage, the 20-40 mesh bagasse was kept at room temperature in an airtight plastic container (Sujan et al. 2018). Moisture and Ash content Moisture content of raw materials was measured using the ASTM D 4442-07 procedure. In a glass crucible, 2 grams of pretreatment sample containing 20-40 mesh were dried in oven at 105±2°C. The moisture content was expressed in percent wet basis and weight measurements were taken every 3 h. A muffle furnace was used for burning the Oven-dried samples at 575±25°C for ash content determination follow- ing the ASTM Standard E 1755-01. Volatile matter The volatile matter analysis was carried out in accordance with ASTM Standard D 271-48. Four grams of raw materials were heated in a furnace at 950±20°C for seven minutes. The weight loss, excluding the weight of moisture pushed off at 105°C, is then used to calculate volatile matter. Fixed carbon The difference between 100 and the total of volatile matter, moisture, and ash content was used to compute the fixed carbon percentage. Chemical analysis The technical association of the pulp and paper industry (TAPPI) method detects α-cellulose (T 203 cm-99), pentosan (T 223 cm-01), klason lignin (T211 om-83), and acid soluble lignin (T UM 250) on a dry basis. Ultimate analysis Ultimate analysis of samples was done by following the procedure ASTM Standard D 5291-02. Organic elemental analyzer (Flash 2000, Thermo Scientific, USA) was used with a specific condition (Reactor temp. 900°C, He: 250 kPa, O2: 250 kPa, TCD). Low concentration alkali pretreatment Different alkali concentration (0%, 2%, 4%, 6% and 8%), time (60, 90 and 120 min) and temperature (80, 100 and 120oC) were applied on raw material to check alkali effect on SB shown in Table 3. A portion of the ground bagasse 20-40 mesh (10g) was taken into plastic zipper bag (temperature 80oC, figure 2a and 2b) and stainless-steel reactor (tempera- ture 100oC and 120oC, figure 2c and 2d). Different alkali concentration was applied during the pretreatment process (alkali to SB ratio ranging between 1:12). The mixture was heated at a particular temperature such as in a water bath (80oC) and oil bath (100oC and 120oC) for a desired length of time. In the course of pretreatment process the sample was manually mixed 2-3 times to attain proper alkali treatment. The treated SB was placed into a polyester bag to remove excess alkali water by pressing it. After that it was vigorously washed with tap water (repeated for five times) to remove the remaining alkali. Finally the SB was dried in an oven at 65oC for 72 h and stored in a close container at room temperature for further experiment. Regression model Regression model has been developed for prediction of α-cel- lulose, pentosan and lignin in SB. The general form of the model is: y=α+α1x1 + α2x2+........................+αnxn + ε.......................(1) where α is the constant term. αi are the coefficient of variables xi. ε is the random error term which is minimized with Simple Least Squares Regression (SLSR). Regression coefficients of the independent variables namely alkali concentration, temperature and time are estimated by SLSR method for developing regression model to predict α-cellulose, pentosan and lignin. Efficiencies of these models are expressed by coefficient of multiple determination (R2) and Adjusted R2. 2.9 Separate hydrolysis and fermentation (SHF) The Hydrolysis experiment took place in 100 ml conical flux 10ml enzyme solution with 200 mg (2% dry wt.) in citrate buffer (0.05 M, pH 5.0) at 50°C for 48 h. In this case Trichoderma viride was used for hydrolysis. Hydrolysate was then heated for 15 min in a boiling water bath and centrifugation was done to remove solid particles. The supernatant was used for analysis of released sugars as described by Jamal et al. (2011). During fermentation process according to Firoz et al. (2012), 100 ml media was prepared and 0.5 g of commercial yeast Saccharomyces cerevisiae was used as inoculum which showed good performance to converts sugar into bioethanol. This inoculated media mixture was poured in a suitable glass- ware and was kept in a shaking incubator for 48 h. 10 ml of this medium was then added into the flask and it was properly covered with aluminum foil. Then it was placed in the incubator at 30oC for 24 h, 48 h, 72 h and 96 h for fermenta- tion of sugars to bio-ethanol according to Sujan et al. (2018). Samples from hydrolysis and fermentation were performed by HPLC. High performance liquid chromatography (HPLC) In characterize part, concentration of Sugars and ethanol were determined by HPLC (Ultimate 3000, Thermo Scien- tific, USA) method using Hyper Rez XP carbohydrate H+ 8 µm column (100×7.7 mm) equipped with a Refractive Index (Shodex RI-101) detector. The mobile phase was degassed with deionized water with a flow rate of 0.7 ml/min and column temperature was maintained at 70°C. It is possible to measure the total sugar concentration in the hydrolysis liquid fraction by comparing its peak area detected by HPLC with peak area of 1% standard sugar which consists of two sugars namely glucose and xylose (Sujan et al. 2018). The same column which is specialized for fermentation broth analysis is used for ethanol detec- tion. The kinetic parameters of ethanol fermentation were determined as follows (Islam et al. 2019): Crystallinity measurement X-ray diffraction (XRD) was used to determine the crystalline structure of the SB samples using a diffrac- tometer (GBC XRD) and filtered copper K radiation (λ = 0.1542 nm) by a monochromator at 35.50 kV voltage and 28 mA current, with a speed of about 2o/min and scan- ning in the range of 10 - 80ºC. The crystallinity index (CrI) was obtained from the ratio between the intensity of the 002 peak (I002, 2θ = 22.5) and the minimum dip (Iam, 2 θ = 18.5) according to the following equation (Roberta et al., 2012): CrI (%) = [(I002 - Iam)/I002] ×100 ...................................... (2) where I002 is the highest peak intensity of plane 002 and Iam is related to the amorphous structure. In present study, the average crystallite sizes were deter- mined from the Scherrer equation by using the diffraction pattern obtained from the 002 (hkl) lattice planes of cellulose samples D(hkl) = [(Kλ / B(hkl) cos2θ] ............................................... (3) Where D(hkl) (Crystallite size), K (Scherrer constant, 0.84), λ (X-ray wavelength, 0.154nm), B(hkl)(Full width half maxi- mum of the measured hkl reflection), and 2θ (Corresponding Bragg angle). Scanning electron microscopy (SEM) analysis In this research, SEM (ZEISS EVO 18 SEM) was used to detect the change of pretreated bagasse fibers. SEM images were taken of different pretreated bagasse samples with acceleration voltage of 2.0 KV. Results and discussion Proximate analysis Proximate analysis of SB sample (20-40 mesh) are presented in Table I. Primarily this analysis usually evaluate the fuel characteristics of raw materials. According to Sun et al., 2009, higher moisture and ashcontent in samples lessen the heating value. Ultimate analysis Ultimate analysis denotes the elemental configuration of SB such as carbon, hydrogen, oxygen, nitrogen and sulfur which are shown in Table II. This examination helps to measure the percentage of carbon and hydrogen content in biomass that is responsible to determine the amount of air is required for complete combustion, composition of combustion gases and heat is generated by it (Poddar et al. 2014). Chemical properties Raw SB contained 34.66±12% α-cellulose, 22.43±08% pentosan and 19.57±06% klason lignin in which 1.75±04% acid soluble lignin (dry basis) was detected by technical association of the pulp and paper industry (TAPPI) method. The chemical composition of SB was determined by acid hydrolysis and it was calculatedby HPLC method as 45.35% glucose and 30.64% xylose (Sujan et al., 2018). α-cellulose yield Bagasse is mainly composed of cellulose, hemicellulose and lignin. Besides these there are some extractives such as ash, wax, gum, pectin etc. During alkali pretreatment usually most of the extractives are removed with the increasing of alkali concentration. Pretreatment of SB with different alkali concentration based on raw material (0%, 2%, 4%, 6% and 8%), time (60, 90 and 120 min) and temperature (80, 100 and 120oC) are shown in Table III. Consequences of each independent experiment varying with alkali concentration, temperature and time on the α-cellulose yield were analyzed using MATLAB software. Apparently, it was observed that α-cellulose yield was considerably increased with changes of alkali concentration ranging from 0% to 8% (34.66% to 63.58%). But no noticeable variation was observed in temperature-time alteration during pretreatment of SB. Based on cellulose percentage obtained in treated bagasse, the optimum conditions for pretreatment reaction were selected as alkali concentration 8%, time 90 min and temperature 100oC. Although cellulose content in treated SB was slowly increased with the increase of alkali concentration 12% attime 90 min and temperature 100oC but as a consequence of huge chemical consumption, recovery problem, chance of losses cellulose and hemicellulose, alkali concentrations for pretreatment above 8% was not considered as ideal concentration. Effect of alkali concentration (AC), cooking time and temperature (temp) charge on α-cellulose yield as well as their statistical significance on the basis of F-test number are presented in regression equations (4). As shown in equations, cooking time at the maximum temperature had no significant effect on α-cellulose yield followed by alkali concentration charge. Effect of temperature on α-cellulose yield was less in employed cooking conditions. For α-cellulose yield: α-cellulose yield = 82.05-0.151×temp-1.56×time-0.69×AC (R2=0.89, adjusted R2=0.87) ............................................. (4) For Pentosan: Pentosan = 13.58+0.086 × tem p + 0.006 × time+0.445 × AC (R2=0.67, adjusted R2=0.63) ............................................. (5) For Lignin: Lignin = 27.662-0.041×temp-0.001×time-01.242×AC (R2=0.93, adjusted R2=0.92) ............................................. (6) For predicting α-cellulose, percentage of pentosan and lignin, the most influential factor was alkali concentration and then cooking temperature for α-cellulose yield, which exhibited an almost linear dependence on both operational variables. The coefficient of determinations is good for α-cel- lulose yield and lignin percentage which hovers around 90 percent, although the figure is moderate more than 60 percent for pentosan. All these three models are significant (p<0.05) at 5% level of significance. In order to perceive the impact of alkali concentration and temperature on these three parameters, three-dimensional (3D) response surface plots were created by plotting the response (α-cellulose yield) pentosan and lignin on the Z-axis versus the most influential one independent variable alkali concentration and temperature as shown in Fig. 3 (a), (b) and (c). XRD analysis In 19th century the cellulose crystalline structure has been discovered and later it was verified by X-ray crystallography (Meyer and Misch, 1937; Wilkie, 1961). The crystallinity index (CrI) of non-woody biomass, such as grasses and agricultural residues, varies depending on the type of biomass. Generally, non-woody biomass has a lower crystallinity index than woody biomass, typically ranging from 20-40%. The crystallinity index of non-woody biomass affects its digestibility and energy production potential.Recently, researchers are being paid more atten- tion on cellulose index because of its potential use in bioen- ergy production. Since then several different models of cellulose index have been proposed. The most popular two-phase cellulose model describes cellulose chains as containing both crystalline (ordered) and amorphous (less ordered) region (Park et al., 2010). Alkali treatment of bagasse has been found to increase the crystallinity index of the bagasse. Alkali treatment breaks down hemicellulose, making the cellulose molecules more ordered and crystalline, resulting in a higher crystallinity index. Alkali treatment also increases the digestibility of the bagasse, making it easier to break down and therefore more suitable for bioenergy production. X-ray diffraction spectra of the SB and alkali treated SB were presented in Figure 4. It was observed that the intensity of 101 and 002 peaks were gradually increased. The relative amount of crystalline cellu- lose (CrI) in the total solid were calculated based on the equation (2) and obtained 18.68%, 24.80%, 27.23%, 32.69%, 36.45% and 39.19% for raw SB, 0%, 2%, 4%, 6% and 8% alkali treated SB samples respectively. The CrI of alkali treated SB samples (not cellulose crystallinity) were intense- ly influenced by the composition of the samples. In case of lignocellulosic biomass examples, cellulose CrI measured the relative amount of crystalline cellulose in the total solid. Therefore, amorphous part of lignin and hemicellulose in biomass specimens were partially removed with the delig- nification process as a result the proportion of α-cellulose was increased and hence CrI would be increased gradually. This interpretation could be proved by the fact that alkali-treated SB had higher CrI than raw SB (Zhao et al. 2010) and the portion of cellulose in the treated SB was also increased gradually. According to the equation 3, the average sizes of crystallite obtained were 3.28, 3.51, 3.63, 3.78, 3.90 and 4.06 nm for raw SB, 0%, 2%, 4%, 6% and 8% alkali treated SB samples respectively. The experimental data revealed that during delignification, the size of crystallite was increased. SEM analysis SEM is one of the most commanding tools widely used to inves- tigate the lignocellulosic biomass surface (Amiri and Karimi, 2015). SEM is usually employed for surface characterization, morphology and inspection of microstructure. In respect of biomass example through SEM images we can compare the untreated and pretreated models which may lead to different insight into the biomass (Karimi and Taherzadeh, 2016). SEM provides two-dimensional images of raw SB and alkali treated SB, which were taken and compared to the outcome of NaOH treatment. All testers were coated with carbon tape and magnification of 500x was used. Figure 4 shows the wall of raw SB (Figure 6a) was intact where the alkali treated SB (Figures 6c-6f), the cell wall was ruptured or splitted and hence packing of the fibers were partially loosened (Firoz et al. 2012). Sugar and ethanol yield Enzymatic digestibility is the ability of enzymes to break down molecules into smaller molecules, such as glucose. It is used to measure the efficiency of enzyme-catalyzed reactions and the digestibility of cellulose. The enzymatic digestibility of a compound is affected by factors such as the compound's crystallinity index, the type of enzyme used, and the temperature and pH of the reaction. In this study the enzymatic digestibility of the alkali pretreated SB was improved by increasing pretreatment conditions. Typically the pretreatment settings are selected by considering various factors such as feedstock characteristics, pretreatment chemical cost, energy consumption and recovery efficiency (Wu et al. 2011). α-cellulose obtained by 8% alkali treated of SB was used for hydrolysis reaction and the reaction was carried out at an optimum condition set by Sujan et al. (2018). The most effective enzymatic hydrolysis was taken place with Trichoderma virideat 48 h and the theoretical yield of sugars i.e. glucose and xylose were obtained 89.59% and 61.23%, respectively. Through fermentation process generated sugars were used to check ethanol production. Yeast Saccharomyces cerevisiae presented worthy performance to convert C6 sugar into ethanol when it was incubated at 30°C for 24 h and 16.81±0.32% ethanol yield was detected. Conclusion α-cellulose yield was optimized by varying alkali concentration, temperature and cooking time. The most vital influencing factor for α-cellulose yield was alkali concentration and after that temperature performed a little bit. The optimum α-cellulose yield (63.58±0.05%) was obtained at 8% alkali concentration, 100oC and 90 minutes. In hydrolysis step of SB, Trichoderma viridaewas used to convert 8% alkali treated SB into sugars and it was attained 89.59% glucose and 61.23% xylose which gave higher yield in compare with our previous study such as ball milled and mesh size varied SB sample (Sujan et al. 2018). At fermentation step Saccha- romyces cerevisiae was used to convert hydrolysate of 8% alkali treated SB sample and ethanol yield was obtained 16.81±0.32% at 24 h. Compositional analysis, imaging and crystallinity are three methods were performed. SEM images can give different clues about SB including morphology, surface disruption and creation of highly accessible surface area iii) CrI (18.68% to 39.19%) and crystal size (3.28 nm to 4.06 nm) of SB are increased with different alkali treatment (0%-8%). In this case it was observed that crystallinity, crystal size, accessible surface area, porosity, particle size, lignin and hemicellulose content and enzyme adsorption/de- sorption were acted as the most impressive factors for digest- ibility of sugarcane bagasse. Acknowledgement Authors would like to acknowledge Pulp and Paper Division, BCSIR for their cooperation. Authors also like to thanks for the sincere assistance of Md. Abdul Hamid, Junior Techni- cian, IFRD, BCSIR. References Alvira P, Tomás-Pejó E, Ballesteros M and Negro MJ (2010), Pretreatment technologies for an efficient bioethanol production process based on enzymatic hydrolysis: a review, Bioresource technology 101(13): 4851-4861. DOI: org/10.1016/j.biortech.2009.11.093 Agbor VB, Cicek N, Sparling R, Berlin A and Levin, DB (2011), Biomass pretreatment: fundamentals toward application, Biotechnology advances 29(6): 675-685. DOI: org/10.1016/j.biotechadv.2011.05.005 Amiri H and Karimi K (2015), Improvement of acetone, butanol, and ethanol production from woody biomass using organosolv pretreatment, Bioprocess and biosys- tems engineering 38(10): 1959-1972. ASTM (2007), Standard test members for instrumental deter- mination of C, H, N in petroleum products and lubricants. ASTM D5291-02, ASTM international, West Conshohocken, PA. Canilha L, Chandel AK, Suzane dos Santos Milessi T, Antunes FAF, Luiz da Costa Freitas W, das Graças Almeida Felipe M and da Silva SS (2012), Bioconver- sion of sugarcane biomass into ethanol: an overview about composition, pretreatment methods, detoxifica- tion of hydrolysates, enzymatic saccharification and ethanol fermentation. Journal of Biomedicine and Biotechnology, DOI: org/10.1155/2012/989572 Den W, Sharma VK, Lee M, Nadadur G and Varma RS (2018), Lignocellulosic biomass transformations via greener oxidative pretreatment processes: access to energy and value-added chemicals, Frontiers in chem- istry 6: 141. DOI: org/10.3389/fchem.2018.00141 Ahmed FM, Rahman SR and Gomes DJ (2012), Saccharifi- cation of sugarcane bagasse by enzymatic treatment for bioethanol production, Malays J Microbiol 8(2), 97-103. Hirani AH, Javed N, Asif M, Basu SK and Kumar A (2018), A review on first-and second-generation biofuel productions. In Biofuels: greenhouse gas mitigation and global warming, Springer, New Delhi, pp 141-154. DOI: 10.1007/978-81-322-3763-1_8 Islam MZ, Asad MA, Hossain MT, Paul SC and Sujan SMA (2019), Bioethanol production from banana pseudostem by using separate and cocultures of cellu- lase enzyme with Saccharomyces cerevisiae, Journal of Environmental Science and Technology 12(4): 157-163. Jahan MS, Saeed A, Ni Y and He Z (2009), Pre-extraction and its impact on the alkaline pulping of bagasse, Journal of Biobased Materials and Bioenergy 3(4): 380-385. DOI: org/10.1166/jbmb.2009.1053 Janker-Obermeier I, Sieber V, Faulstich Mand Schieder D (2012), Solubilization of hemicellulose and lignin from wheat straw through microwave-assisted alkali treatment, Industrial Crops and Products 39: 198-203. DOI: org/10.1016/j.indcrop.2012.02.022 Karimi K and Taherzadeh MJ (2016), A critical review of analytical methods in pretreatment of lignocelluloses: composition, imaging, and crystallinity, Bioresource technology 200: 1008-1018. DOI: org/10.1016/j. biortech.2015.11.022 Maryana R, Ma’rifatun D, Wheni AI, Satriyo KW and Rizal WA (2014), Alkaline pretreatment on sugarcane bagasse for bioethanol production, Energy Procedia, 47: 250-254. DOI: org/10.1016/j.egypro.2014.01.221 McIntosh Sand Vancov T (2010), Enhanced enzyme sacchar- ification of Sorghum bicolor straw using dilute alkali pretreatment, Bioresource technology 101(17): 6718-6727. DOI: org/10.1016/j.biortech.2010.03.116 Meyer KH and Misch L (1937), Positions des atomes dans le nouveau modele spatial de la cellulose. Helvetica Chimica Acta 20(1): 232-244. DOI: org/10. 1002/hlca.19370200134 Park S, Baker JO, Himmel ME, Parilla PA and Johnson DK (2010), Cellulose crystallinity index: measurement techniques and their impact on interpreting cellulase performance, Biotechnology for biofuels 3(1): 10. Sujan SMA, Bari ML and Fakhruddin AN (2018), Effects of physical pretreatment (crushing and ball milling) on sugarcane bagasse for bioethanol production, Bangla- desh journal of botany 147(2): 257-64. Wilkie JS (1961), Carl Nägeli and the fine structure of living matter, Nature 190: 1145-1150. Wu L, Arakane M, Ike M, Wada M, Takai T, Gau Mand Tokuyasu K (2011), Low temperature alkali pretreat- ment for improving enzymatic digestibility of sweet sorghum bagasse for ethanol production, Bioresource Technology 102(7): 4793-4799. Zhang W, Okubayashi S and Bechtold T (2005), Fibrillation tendency of cellulosic fibers-Part 4. Effects of alkali pretreatment of various cellulosic fibers, Carbohydrate polymers. 61(4): 427-433. Zhao X, van der Heide E, Zhang Tand Liu D (2010), Delig- nification of sugarcane bagasse with alkali and peracetic acid and characterization of the pulp, BioRe- sources 5(3): 1565-1580. Zhao X, Zhang Land Liu D (2012), Biomass recalcitrance. Part II: Fundamentals of different pre‐treatments to increase the enzymatic digestibility of lignocellulose, Biofuels, Bioproducts and Biorefining 6(5): 561-579. Table II. Ultimate analysis of Sugarcane bagasse (20-40 mesh) Name of sample Carbon (%) Oxygen (%) Hydrogen (%) Nitrogen (%) Sulfur (%) SB 44.89 49.6 5.51 0 0 Table III. Chemical pretreatment of SB with different concentration of NaOH, temperature and time for the yield of α-cellulose, pentosan and lignin yield Alkali (NaOH) solution (%) Temperature (oC) Time (min) α-cellulose yield (%) Pentosan yield (%) Lignin yield (%) 0% 80 60 34.66±0.05 22.43±0.02 19.57±0.03 0% 80 90 34.76±0.02 22.56±0.01 19.42±0.01 0% 100 60 34.88±0.01 22.68±0.03 19.39±0.01 0% 100 120 34.90±0.01 22.75±0.02 19.25±0.06 0% 120 90 34.98±0.02 22.88±0.02 19.16±0.01 2% 80 60 37.02±0.10 23.20±0.01 17.77±0.09 2% 80 90 37.32±0.08 23.35±0.03 17.23±0.05 2% 100 60 38.13±0.03 23.60±0.04 17.14±0.01 2% 100 90 38.99±0.05 23.84±0.03 17.05±0.02 2% 120 120 38.46±0.08 23.25±0.02 16.73±0.06 4% 80 90 44.98±0.13 23.96±0.05 14.96±0.08 4% 80 120 45.46±0.09 24.26±0.08 14.85±0.01 4% 100 60 47.43±0.11 24.53±0.04 14.66±0.05 4% 100 90 48.05±0.06 24.87±0.05 14.50±0.01 4% 120 60 47.86±0.09 24.95±0.02 14.41±0.06 6% 80 60 53.45±0.11 25.09±0.03 12.86±0.08 6% 80 90 53.95±0.08 25.21±0.02 12.58±0.06 6% 100 90 55.79±0.05 25.86±0.03 12.29±0.02 6% 100 120 54.35±0.03 25.89±0.01 12.11±0.02 6% 120 60 54.88±0.04 25.99±0.02 12.01±0.02 8% 80 60 59.24±0.14 26.15±0.04 10.79±0.10 8% 80 120 59.95±0.02 26.32±0.02 10.69±0.01 8% 100 60 61.67±0.11 26.56±0.03 10.17±0.08 8% 100 90 63.58±0.05 26.75±0.02 09.01±0.02 8% 120 90 62.59±0.04 26.86±0.02 09.30±0.04 12% 100 90 66.14±0.06 27.46±0.04 06.22±0.05 Optimization of alkali concentration in the pretreatment of sugarcane bagasse 58(2) 202394 composed of vasculer bundles surrounded by non-fibrous parenchymatic cell (Jahan et al. 2009). Usually it is available as agricultural residue and byproduct from sugar mill indus- tries. In the year of 2015-16, there was 2,95,162 metric ton sugarcane bagasse was obtained from sugar mills of Bangla- desh (http://www.bsfic.gov.bd). To enrich cellulose content in biomass for enzymatic saccarification, current research has focused on various pretreatment processes. Currently, scientists have gone through a lot of studies on different pretreatment techniques to remove the compact and rigid composition by open up the cellulosic structure (Alvira et al. 2010). Pretreatment technologies are commonly abandoned to reduce the structural barriers and boost cellulose availabili- ty based on not only chemicals such as acid, alkali, oxidant, etc. but also several treatment settings. Alkali treatment is one of the most widespread and economical methods used for surface modification of lignocellulosic biomass. In the field of biorefinary, alkali pretreatment is intensively employed to develop the cellulosic materials both mechan- ically as well as chemically and such properties include tensile strength, dyeability, stability of dimension and reactivity (Wu et al. 2011). In alkali pretreatment biomass is treated under moderate reaction conditions ensuring inexpensiveness, inflated recycling possibilities of water and chemical agents (McIntosh and Vancov, 2010). Usually alkali treatment is the most effective pretreatment for agricultural residues, herbaceous crops and hardwood containing low lignin content in the compariosn of softwood containing high lignin content (Agbor et al. 2011; Canilha et al. 2012). This research aims is to investigate the alkaline pretreatment repercussions based on the chemical structure, surface morphology, structure and enzymatic digestion of sugarcane bagasse for sugars and ethanol production. Materials and methods Sugarcane bagasse (SB) was taken from street juice vendor. Warm tap water was used for washing purpose to eliminate the residual free sugars once again. This washing and pressing step was then repeated for three times. After that drying was done in an oven at 65oC for 16 h. Crushing of dried bagasse was done successively using a locally made crusher and sieved (Retsch, D-42759, HAAN, Germany) to have the particle size of 20-40 mesh on average. Before usage, the 20-40 mesh bagasse was kept at room temperature in an airtight plastic container (Sujan et al. 2018). Moisture and Ash content Moisture content of raw materials was measured using the ASTM D 4442-07 procedure. In a glass crucible, 2 grams of pretreatment sample containing 20-40 mesh were dried in oven at 105±2°C. The moisture content was expressed in percent wet basis and weight measurements were taken every 3 h. A muffle furnace was used for burning the Oven-dried samples at 575±25°C for ash content determination follow- ing the ASTM Standard E 1755-01. Volatile matter The volatile matter analysis was carried out in accordance with ASTM Standard D 271-48. Four grams of raw materials were heated in a furnace at 950±20°C for seven minutes. The weight loss, excluding the weight of moisture pushed off at 105°C, is then used to calculate volatile matter. Fixed carbon The difference between 100 and the total of volatile matter, moisture, and ash content was used to compute the fixed carbon percentage. Chemical analysis The technical association of the pulp and paper industry (TAPPI) method detects α-cellulose (T 203 cm-99), pentosan (T 223 cm-01), klason lignin (T211 om-83), and acid soluble lignin (T UM 250) on a dry basis. Ultimate analysis Ultimate analysis of samples was done by following the procedure ASTM Standard D 5291-02. Organic elemental analyzer (Flash 2000, Thermo Scientific, USA) was used with a specific condition (Reactor temp. 900°C, He: 250 kPa, O2: 250 kPa, TCD). Low concentration alkali pretreatment Different alkali concentration (0%, 2%, 4%, 6% and 8%), time (60, 90 and 120 min) and temperature (80, 100 and 120oC) were applied on raw material to check alkali effect on SB shown in Table 3. A portion of the ground bagasse 20-40 mesh (10g) was taken into plastic zipper bag (temperature 80oC, figure 2a and 2b) and stainless-steel reactor (tempera- ture 100oC and 120oC, figure 2c and 2d). Different alkali concentration was applied during the pretreatment process (alkali to SB ratio ranging between 1:12). The mixture was heated at a particular temperature such as in a water bath (80oC) and oil bath (100oC and 120oC) for a desired length of time. In the course of pretreatment process the sample was manually mixed 2-3 times to attain proper alkali treatment. The treated SB was placed into a polyester bag to remove excess alkali water by pressing it. After that it was vigorously washed with tap water (repeated for five times) to remove the remaining alkali. Finally the SB was dried in an oven at 65oC for 72 h and stored in a close container at room temperature for further experiment. Regression model Regression model has been developed for prediction of α-cel- lulose, pentosan and lignin in SB. The general form of the model is: y=α+α1x1 + α2x2+........................+αnxn + ε.......................(1) where α is the constant term. αi are the coefficient of variables xi. ε is the random error term which is minimized with Simple Least Squares Regression (SLSR). Regression coefficients of the independent variables namely alkali concentration, temperature and time are estimated by SLSR method for developing regression model to predict α-cellulose, pentosan and lignin. Efficiencies of these models are expressed by coefficient of multiple determination (R2) and Adjusted R2. 2.9 Separate hydrolysis and fermentation (SHF) The Hydrolysis experiment took place in 100 ml conical flux 10ml enzyme solution with 200 mg (2% dry wt.) in citrate buffer (0.05 M, pH 5.0) at 50°C for 48 h. In this case Trichoderma viride was used for hydrolysis. Hydrolysate was then heated for 15 min in a boiling water bath and centrifugation was done to remove solid particles. The supernatant was used for analysis of released sugars as described by Jamal et al. (2011). During fermentation process according to Firoz et al. (2012), 100 ml media was prepared and 0.5 g of commercial yeast Saccharomyces cerevisiae was used as inoculum which showed good performance to converts sugar into bioethanol. This inoculated media mixture was poured in a suitable glass- ware and was kept in a shaking incubator for 48 h. 10 ml of this medium was then added into the flask and it was properly covered with aluminum foil. Then it was placed in the incubator at 30oC for 24 h, 48 h, 72 h and 96 h for fermenta- tion of sugars to bio-ethanol according to Sujan et al. (2018). Samples from hydrolysis and fermentation were performed by HPLC. High performance liquid chromatography (HPLC) In characterize part, concentration of Sugars and ethanol were determined by HPLC (Ultimate 3000, Thermo Scien- tific, USA) method using Hyper Rez XP carbohydrate H+ 8 µm column (100×7.7 mm) equipped with a Refractive Index (Shodex RI-101) detector. The mobile phase was degassed with deionized water with a flow rate of 0.7 ml/min and column temperature was maintained at 70°C. It is possible to measure the total sugar concentration in the hydrolysis liquid fraction by comparing its peak area detected by HPLC with peak area of 1% standard sugar which consists of two sugars namely glucose and xylose (Sujan et al. 2018). The same column which is specialized for fermentation broth analysis is used for ethanol detec- tion. The kinetic parameters of ethanol fermentation were determined as follows (Islam et al. 2019): Crystallinity measurement X-ray diffraction (XRD) was used to determine the crystalline structure of the SB samples using a diffrac- tometer (GBC XRD) and filtered copper K radiation (λ = 0.1542 nm) by a monochromator at 35.50 kV voltage and 28 mA current, with a speed of about 2o/min and scan- ning in the range of 10 - 80ºC. The crystallinity index (CrI) was obtained from the ratio between the intensity of the 002 peak (I002, 2θ = 22.5) and the minimum dip (Iam, 2 θ = 18.5) according to the following equation (Roberta et al., 2012): CrI (%) = [(I002 - Iam)/I002] ×100 ...................................... (2) where I002 is the highest peak intensity of plane 002 and Iam is related to the amorphous structure. In present study, the average crystallite sizes were deter- mined from the Scherrer equation by using the diffraction pattern obtained from the 002 (hkl) lattice planes of cellulose samples D(hkl) = [(Kλ / B(hkl) cos2θ] ............................................... (3) Where D(hkl) (Crystallite size), K (Scherrer constant, 0.84), λ (X-ray wavelength, 0.154nm), B(hkl)(Full width half maxi- mum of the measured hkl reflection), and 2θ (Corresponding Bragg angle). Scanning electron microscopy (SEM) analysis In this research, SEM (ZEISS EVO 18 SEM) was used to detect the change of pretreated bagasse fibers. SEM images were taken of different pretreated bagasse samples with acceleration voltage of 2.0 KV. Results and discussion Proximate analysis Proximate analysis of SB sample (20-40 mesh) are presented in Table I. Primarily this analysis usually evaluate the fuel characteristics of raw materials. According to Sun et al., 2009, higher moisture and ashcontent in samples lessen the heating value. Ultimate analysis Ultimate analysis denotes the elemental configuration of SB such as carbon, hydrogen, oxygen, nitrogen and sulfur which are shown in Table II. This examination helps to measure the percentage of carbon and hydrogen content in biomass that is responsible to determine the amount of air is required for complete combustion, composition of combustion gases and heat is generated by it (Poddar et al. 2014). Chemical properties Raw SB contained 34.66±12% α-cellulose, 22.43±08% pentosan and 19.57±06% klason lignin in which 1.75±04% acid soluble lignin (dry basis) was detected by technical association of the pulp and paper industry (TAPPI) method. The chemical composition of SB was determined by acid hydrolysis and it was calculatedby HPLC method as 45.35% glucose and 30.64% xylose (Sujan et al., 2018). α-cellulose yield Bagasse is mainly composed of cellulose, hemicellulose and lignin. Besides these there are some extractives such as ash, wax, gum, pectin etc. During alkali pretreatment usually most of the extractives are removed with the increasing of alkali concentration. Pretreatment of SB with different alkali concentration based on raw material (0%, 2%, 4%, 6% and 8%), time (60, 90 and 120 min) and temperature (80, 100 and 120oC) are shown in Table III. Consequences of each independent experiment varying with alkali concentration, temperature and time on the α-cellulose yield were analyzed using MATLAB software. Apparently, it was observed that α-cellulose yield was considerably increased with changes of alkali concentration ranging from 0% to 8% (34.66% to 63.58%). But no noticeable variation was observed in temperature-time alteration during pretreatment of SB. Based on cellulose percentage obtained in treated bagasse, the optimum conditions for pretreatment reaction were selected as alkali concentration 8%, time 90 min and temperature 100oC. Although cellulose content in treated SB was slowly increased with the increase of alkali concentration 12% attime 90 min and temperature 100oC but as a consequence of huge chemical consumption, recovery problem, chance of losses cellulose and hemicellulose, alkali concentrations for pretreatment above 8% was not considered as ideal concentration. Effect of alkali concentration (AC), cooking time and temperature (temp) charge on α-cellulose yield as well as their statistical significance on the basis of F-test number are presented in regression equations (4). As shown in equations, cooking time at the maximum temperature had no significant effect on α-cellulose yield followed by alkali concentration charge. Effect of temperature on α-cellulose yield was less in employed cooking conditions. For α-cellulose yield: α-cellulose yield = 82.05-0.151×temp-1.56×time-0.69×AC (R2=0.89, adjusted R2=0.87) ............................................. (4) For Pentosan: Pentosan = 13.58+0.086 × tem p + 0.006 × time+0.445 × AC (R2=0.67, adjusted R2=0.63) ............................................. (5) For Lignin: Lignin = 27.662-0.041×temp-0.001×time-01.242×AC (R2=0.93, adjusted R2=0.92) ............................................. (6) For predicting α-cellulose, percentage of pentosan and lignin, the most influential factor was alkali concentration and then cooking temperature for α-cellulose yield, which exhibited an almost linear dependence on both operational variables. The coefficient of determinations is good for α-cel- lulose yield and lignin percentage which hovers around 90 percent, although the figure is moderate more than 60 percent for pentosan. All these three models are significant (p<0.05) at 5% level of significance. In order to perceive the impact of alkali concentration and temperature on these three parameters, three-dimensional (3D) response surface plots were created by plotting the response (α-cellulose yield) pentosan and lignin on the Z-axis versus the most influential one independent variable alkali concentration and temperature as shown in Fig. 3 (a), (b) and (c). XRD analysis In 19th century the cellulose crystalline structure has been discovered and later it was verified by X-ray crystallography (Meyer and Misch, 1937; Wilkie, 1961). The crystallinity index (CrI) of non-woody biomass, such as grasses and agricultural residues, varies depending on the type of biomass. Generally, non-woody biomass has a lower crystallinity index than woody biomass, typically ranging from 20-40%. The crystallinity index of non-woody biomass affects its digestibility and energy production potential.Recently, researchers are being paid more atten- tion on cellulose index because of its potential use in bioen- ergy production. Since then several different models of cellulose index have been proposed. The most popular two-phase cellulose model describes cellulose chains as containing both crystalline (ordered) and amorphous (less ordered) region (Park et al., 2010). Alkali treatment of bagasse has been found to increase the crystallinity index of the bagasse. Alkali treatment breaks down hemicellulose, making the cellulose molecules more ordered and crystalline, resulting in a higher crystallinity index. Alkali treatment also increases the digestibility of the bagasse, making it easier to break down and therefore more suitable for bioenergy production. X-ray diffraction spectra of the SB and alkali treated SB were presented in Figure 4. It was observed that the intensity of 101 and 002 peaks were gradually increased. The relative amount of crystalline cellu- lose (CrI) in the total solid were calculated based on the equation (2) and obtained 18.68%, 24.80%, 27.23%, 32.69%, 36.45% and 39.19% for raw SB, 0%, 2%, 4%, 6% and 8% alkali treated SB samples respectively. The CrI of alkali treated SB samples (not cellulose crystallinity) were intense- ly influenced by the composition of the samples. In case of lignocellulosic biomass examples, cellulose CrI measured the relative amount of crystalline cellulose in the total solid. Therefore, amorphous part of lignin and hemicellulose in biomass specimens were partially removed with the delig- nification process as a result the proportion of α-cellulose was increased and hence CrI would be increased gradually. This interpretation could be proved by the fact that alkali-treated SB had higher CrI than raw SB (Zhao et al. 2010) and the portion of cellulose in the treated SB was also increased gradually. According to the equation 3, the average sizes of crystallite obtained were 3.28, 3.51, 3.63, 3.78, 3.90 and 4.06 nm for raw SB, 0%, 2%, 4%, 6% and 8% alkali treated SB samples respectively. The experimental data revealed that during delignification, the size of crystallite was increased. SEM analysis SEM is one of the most commanding tools widely used to inves- tigate the lignocellulosic biomass surface (Amiri and Karimi, 2015). SEM is usually employed for surface characterization, morphology and inspection of microstructure. In respect of biomass example through SEM images we can compare the untreated and pretreated models which may lead to different insight into the biomass (Karimi and Taherzadeh, 2016). SEM provides two-dimensional images of raw SB and alkali treated SB, which were taken and compared to the outcome of NaOH treatment. All testers were coated with carbon tape and magnification of 500x was used. Figure 4 shows the wall of raw SB (Figure 6a) was intact where the alkali treated SB (Figures 6c-6f), the cell wall was ruptured or splitted and hence packing of the fibers were partially loosened (Firoz et al. 2012). Sugar and ethanol yield Enzymatic digestibility is the ability of enzymes to break down molecules into smaller molecules, such as glucose. It is used to measure the efficiency of enzyme-catalyzed reactions and the digestibility of cellulose. The enzymatic digestibility of a compound is affected by factors such as the compound's crystallinity index, the type of enzyme used, and the temperature and pH of the reaction. In this study the enzymatic digestibility of the alkali pretreated SB was improved by increasing pretreatment conditions. Typically the pretreatment settings are selected by considering various factors such as feedstock characteristics, pretreatment chemical cost, energy consumption and recovery efficiency (Wu et al. 2011). α-cellulose obtained by 8% alkali treated of SB was used for hydrolysis reaction and the reaction was carried out at an optimum condition set by Sujan et al. (2018). The most effective enzymatic hydrolysis was taken place with Trichoderma virideat 48 h and the theoretical yield of sugars i.e. glucose and xylose were obtained 89.59% and 61.23%, respectively. Through fermentation process generated sugars were used to check ethanol production. Yeast Saccharomyces cerevisiae presented worthy performance to convert C6 sugar into ethanol when it was incubated at 30°C for 24 h and 16.81±0.32% ethanol yield was detected. Conclusion α-cellulose yield was optimized by varying alkali concentration, temperature and cooking time. The most vital influencing factor for α-cellulose yield was alkali concentration and after that temperature performed a little bit. The optimum α-cellulose yield (63.58±0.05%) was obtained at 8% alkali concentration, 100oC and 90 minutes. In hydrolysis step of SB, Trichoderma viridaewas used to convert 8% alkali treated SB into sugars and it was attained 89.59% glucose and 61.23% xylose which gave higher yield in compare with our previous study such as ball milled and mesh size varied SB sample (Sujan et al. 2018). At fermentation step Saccha- romyces cerevisiae was used to convert hydrolysate of 8% alkali treated SB sample and ethanol yield was obtained 16.81±0.32% at 24 h. Compositional analysis, imaging and crystallinity are three methods were performed. SEM images can give different clues about SB including morphology, surface disruption and creation of highly accessible surface area iii) CrI (18.68% to 39.19%) and crystal size (3.28 nm to 4.06 nm) of SB are increased with different alkali treatment (0%-8%). In this case it was observed that crystallinity, crystal size, accessible surface area, porosity, particle size, lignin and hemicellulose content and enzyme adsorption/de- sorption were acted as the most impressive factors for digest- ibility of sugarcane bagasse. Acknowledgement Authors would like to acknowledge Pulp and Paper Division, BCSIR for their cooperation. Authors also like to thanks for the sincere assistance of Md. Abdul Hamid, Junior Techni- cian, IFRD, BCSIR. References Alvira P, Tomás-Pejó E, Ballesteros M and Negro MJ (2010), Pretreatment technologies for an efficient bioethanol production process based on enzymatic hydrolysis: a review, Bioresource technology 101(13): 4851-4861. DOI: org/10.1016/j.biortech.2009.11.093 Agbor VB, Cicek N, Sparling R, Berlin A and Levin, DB (2011), Biomass pretreatment: fundamentals toward application, Biotechnology advances 29(6): 675-685. DOI: org/10.1016/j.biotechadv.2011.05.005 Amiri H and Karimi K (2015), Improvement of acetone, butanol, and ethanol production from woody biomass using organosolv pretreatment, Bioprocess and biosys- tems engineering 38(10): 1959-1972. ASTM (2007), Standard test members for instrumental deter- mination of C, H, N in petroleum products and lubricants. ASTM D5291-02, ASTM international, West Conshohocken, PA. Canilha L, Chandel AK, Suzane dos Santos Milessi T, Antunes FAF, Luiz da Costa Freitas W, das Graças Almeida Felipe M and da Silva SS (2012), Bioconver- sion of sugarcane biomass into ethanol: an overview about composition, pretreatment methods, detoxifica- tion of hydrolysates, enzymatic saccharification and ethanol fermentation. Journal of Biomedicine and Biotechnology, DOI: org/10.1155/2012/989572 Den W, Sharma VK, Lee M, Nadadur G and Varma RS (2018), Lignocellulosic biomass transformations via greener oxidative pretreatment processes: access to energy and value-added chemicals, Frontiers in chem- istry 6: 141. DOI: org/10.3389/fchem.2018.00141 Ahmed FM, Rahman SR and Gomes DJ (2012), Saccharifi- cation of sugarcane bagasse by enzymatic treatment for bioethanol production, Malays J Microbiol 8(2), 97-103. Hirani AH, Javed N, Asif M, Basu SK and Kumar A (2018), A review on first-and second-generation biofuel productions. In Biofuels: greenhouse gas mitigation and global warming, Springer, New Delhi, pp 141-154. DOI: 10.1007/978-81-322-3763-1_8 Islam MZ, Asad MA, Hossain MT, Paul SC and Sujan SMA (2019), Bioethanol production from banana pseudostem by using separate and cocultures of cellu- lase enzyme with Saccharomyces cerevisiae, Journal of Environmental Science and Technology 12(4): 157-163. Jahan MS, Saeed A, Ni Y and He Z (2009), Pre-extraction and its impact on the alkaline pulping of bagasse, Journal of Biobased Materials and Bioenergy 3(4): 380-385. DOI: org/10.1166/jbmb.2009.1053 Janker-Obermeier I, Sieber V, Faulstich Mand Schieder D (2012), Solubilization of hemicellulose and lignin from wheat straw through microwave-assisted alkali treatment, Industrial Crops and Products 39: 198-203. DOI: org/10.1016/j.indcrop.2012.02.022 Karimi K and Taherzadeh MJ (2016), A critical review of analytical methods in pretreatment of lignocelluloses: composition, imaging, and crystallinity, Bioresource technology 200: 1008-1018. DOI: org/10.1016/j. biortech.2015.11.022 Maryana R, Ma’rifatun D, Wheni AI, Satriyo KW and Rizal WA (2014), Alkaline pretreatment on sugarcane bagasse for bioethanol production, Energy Procedia, 47: 250-254. DOI: org/10.1016/j.egypro.2014.01.221 McIntosh Sand Vancov T (2010), Enhanced enzyme sacchar- ification of Sorghum bicolor straw using dilute alkali pretreatment, Bioresource technology 101(17): 6718-6727. DOI: org/10.1016/j.biortech.2010.03.116 Meyer KH and Misch L (1937), Positions des atomes dans le nouveau modele spatial de la cellulose. Helvetica Chimica Acta 20(1): 232-244. DOI: org/10. 1002/hlca.19370200134 Park S, Baker JO, Himmel ME, Parilla PA and Johnson DK (2010), Cellulose crystallinity index: measurement techniques and their impact on interpreting cellulase performance, Biotechnology for biofuels 3(1): 10. Sujan SMA, Bari ML and Fakhruddin AN (2018), Effects of physical pretreatment (crushing and ball milling) on sugarcane bagasse for bioethanol production, Bangla- desh journal of botany 147(2): 257-64. Wilkie JS (1961), Carl Nägeli and the fine structure of living matter, Nature 190: 1145-1150. Wu L, Arakane M, Ike M, Wada M, Takai T, Gau Mand Tokuyasu K (2011), Low temperature alkali pretreat- ment for improving enzymatic digestibility of sweet sorghum bagasse for ethanol production, Bioresource Technology 102(7): 4793-4799. Zhang W, Okubayashi S and Bechtold T (2005), Fibrillation tendency of cellulosic fibers-Part 4. Effects of alkali pretreatment of various cellulosic fibers, Carbohydrate polymers. 61(4): 427-433. Zhao X, van der Heide E, Zhang Tand Liu D (2010), Delig- nification of sugarcane bagasse with alkali and peracetic acid and characterization of the pulp, BioRe- sources 5(3): 1565-1580. Zhao X, Zhang Land Liu D (2012), Biomass recalcitrance. Part II: Fundamentals of different pre‐treatments to increase the enzymatic digestibility of lignocellulose, Biofuels, Bioproducts and Biorefining 6(5): 561-579. Fig. 3. Surface plot of (a) α-cellulose, (b) Pentosan and (c) Lignin against alkali concentration and temperature 120 A lfa cellulose 30 40 100 50 60 Temperature0.0 2.5 805.0 7.5 A lk ali concentration 120 Pentosan 20 100 22 24 26 T emperature 0.0 2.5 805.0 7.5 A lk ali concentr ation 120 Lignin 10 100 15 20 T emper ature 25 0.0 2.5 805.0 7.5 A lk ali concentration 3a 3b 3c Sujan, Hossain, Uddin and Fakhruddin 95 composed of vasculer bundles surrounded by non-fibrous parenchymatic cell (Jahan et al. 2009). Usually it is available as agricultural residue and byproduct from sugar mill indus- tries. In the year of 2015-16, there was 2,95,162 metric ton sugarcane bagasse was obtained from sugar mills of Bangla- desh (http://www.bsfic.gov.bd). To enrich cellulose content in biomass for enzymatic saccarification, current research has focused on various pretreatment processes. Currently, scientists have gone through a lot of studies on different pretreatment techniques to remove the compact and rigid composition by open up the cellulosic structure (Alvira et al. 2010). Pretreatment technologies are commonly abandoned to reduce the structural barriers and boost cellulose availabili- ty based on not only chemicals such as acid, alkali, oxidant, etc. but also several treatment settings. Alkali treatment is one of the most widespread and economical methods used for surface modification of lignocellulosic biomass. In the field of biorefinary, alkali pretreatment is intensively employed to develop the cellulosic materials both mechan- ically as well as chemically and such properties include tensile strength, dyeability, stability of dimension and reactivity (Wu et al. 2011). In alkali pretreatment biomass is treated under moderate reaction conditions ensuring inexpensiveness, inflated recycling possibilities of water and chemical agents (McIntosh and Vancov, 2010). Usually alkali treatment is the most effective pretreatment for agricultural residues, herbaceous crops and hardwood containing low lignin content in the compariosn of softwood containing high lignin content (Agbor et al. 2011; Canilha et al. 2012). This research aims is to investigate the alkaline pretreatment repercussions based on the chemical structure, surface morphology, structure and enzymatic digestion of sugarcane bagasse for sugars and ethanol production. Materials and methods Sugarcane bagasse (SB) was taken from street juice vendor. Warm tap water was used for washing purpose to eliminate the residual free sugars once again. This washing and pressing step was then repeated for three times. After that drying was done in an oven at 65oC for 16 h. Crushing of dried bagasse was done successively using a locally made crusher and sieved (Retsch, D-42759, HAAN, Germany) to have the particle size of 20-40 mesh on average. Before usage, the 20-40 mesh bagasse was kept at room temperature in an airtight plastic container (Sujan et al. 2018). Moisture and Ash content Moisture content of raw materials was measured using the ASTM D 4442-07 procedure. In a glass crucible, 2 grams of pretreatment sample containing 20-40 mesh were dried in oven at 105±2°C. The moisture content was expressed in percent wet basis and weight measurements were taken every 3 h. A muffle furnace was used for burning the Oven-dried samples at 575±25°C for ash content determination follow- ing the ASTM Standard E 1755-01. Volatile matter The volatile matter analysis was carried out in accordance with ASTM Standard D 271-48. Four grams of raw materials were heated in a furnace at 950±20°C for seven minutes. The weight loss, excluding the weight of moisture pushed off at 105°C, is then used to calculate volatile matter. Fixed carbon The difference between 100 and the total of volatile matter, moisture, and ash content was used to compute the fixed carbon percentage. Chemical analysis The technical association of the pulp and paper industry (TAPPI) method detects α-cellulose (T 203 cm-99), pentosan (T 223 cm-01), klason lignin (T211 om-83), and acid soluble lignin (T UM 250) on a dry basis. Ultimate analysis Ultimate analysis of samples was done by following the procedure ASTM Standard D 5291-02. Organic elemental analyzer (Flash 2000, Thermo Scientific, USA) was used with a specific condition (Reactor temp. 900°C, He: 250 kPa, O2: 250 kPa, TCD). Low concentration alkali pretreatment Different alkali concentration (0%, 2%, 4%, 6% and 8%), time (60, 90 and 120 min) and temperature (80, 100 and 120oC) were applied on raw material to check alkali effect on SB shown in Table 3. A portion of the ground bagasse 20-40 mesh (10g) was taken into plastic zipper bag (temperature 80oC, figure 2a and 2b) and stainless-steel reactor (tempera- ture 100oC and 120oC, figure 2c and 2d). Different alkali concentration was applied during the pretreatment process (alkali to SB ratio ranging between 1:12). The mixture was heated at a particular temperature such as in a water bath (80oC) and oil bath (100oC and 120oC) for a desired length of time. In the course of pretreatment process the sample was manually mixed 2-3 times to attain proper alkali treatment. The treated SB was placed into a polyester bag to remove excess alkali water by pressing it. After that it was vigorously washed with tap water (repeated for five times) to remove the remaining alkali. Finally the SB was dried in an oven at 65oC for 72 h and stored in a close container at room temperature for further experiment. Regression model Regression model has been developed for prediction of α-cel- lulose, pentosan and lignin in SB. The general form of the model is: y=α+α1x1 + α2x2+........................+αnxn + ε.......................(1) where α is the constant term. αi are the coefficient of variables xi. ε is the random error term which is minimized with Simple Least Squares Regression (SLSR). Regression coefficients of the independent variables namely alkali concentration, temperature and time are estimated by SLSR method for developing regression model to predict α-cellulose, pentosan and lignin. Efficiencies of these models are expressed by coefficient of multiple determination (R2) and Adjusted R2. 2.9 Separate hydrolysis and fermentation (SHF) The Hydrolysis experiment took place in 100 ml conical flux 10ml enzyme solution with 200 mg (2% dry wt.) in citrate buffer (0.05 M, pH 5.0) at 50°C for 48 h. In this case Trichoderma viride was used for hydrolysis. Hydrolysate was then heated for 15 min in a boiling water bath and centrifugation was done to remove solid particles. The supernatant was used for analysis of released sugars as described by Jamal et al. (2011). During fermentation process according to Firoz et al. (2012), 100 ml media was prepared and 0.5 g of commercial yeast Saccharomyces cerevisiae was used as inoculum which showed good performance to converts sugar into bioethanol. This inoculated media mixture was poured in a suitable glass- ware and was kept in a shaking incubator for 48 h. 10 ml of this medium was then added into the flask and it was properly covered with aluminum foil. Then it was placed in the incubator at 30oC for 24 h, 48 h, 72 h and 96 h for fermenta- tion of sugars to bio-ethanol according to Sujan et al. (2018). Samples from hydrolysis and fermentation were performed by HPLC. High performance liquid chromatography (HPLC) In characterize part, concentration of Sugars and ethanol were determined by HPLC (Ultimate 3000, Thermo Scien- tific, USA) method using Hyper Rez XP carbohydrate H+ 8 µm column (100×7.7 mm) equipped with a Refractive Index (Shodex RI-101) detector. The mobile phase was degassed with deionized water with a flow rate of 0.7 ml/min and column temperature was maintained at 70°C. It is possible to measure the total sugar concentration in the hydrolysis liquid fraction by comparing its peak area detected by HPLC with peak area of 1% standard sugar which consists of two sugars namely glucose and xylose (Sujan et al. 2018). The same column which is specialized for fermentation broth analysis is used for ethanol detec- tion. The kinetic parameters of ethanol fermentation were determined as follows (Islam et al. 2019): Crystallinity measurement X-ray diffraction (XRD) was used to determine the crystalline structure of the SB samples using a diffrac- tometer (GBC XRD) and filtered copper K radiation (λ = 0.1542 nm) by a monochromator at 35.50 kV voltage and 28 mA current, with a speed of about 2o/min and scan- ning in the range of 10 - 80ºC. The crystallinity index (CrI) was obtained from the ratio between the intensity of the 002 peak (I002, 2θ = 22.5) and the minimum dip (Iam, 2 θ = 18.5) according to the following equation (Roberta et al., 2012): CrI (%) = [(I002 - Iam)/I002] ×100 ...................................... (2) where I002 is the highest peak intensity of plane 002 and Iam is related to the amorphous structure. In present study, the average crystallite sizes were deter- mined from the Scherrer equation by using the diffraction pattern obtained from the 002 (hkl) lattice planes of cellulose samples D(hkl) = [(Kλ / B(hkl) cos2θ] ............................................... (3) Where D(hkl) (Crystallite size), K (Scherrer constant, 0.84), λ (X-ray wavelength, 0.154nm), B(hkl)(Full width half maxi- mum of the measured hkl reflection), and 2θ (Corresponding Bragg angle). Scanning electron microscopy (SEM) analysis In this research, SEM (ZEISS EVO 18 SEM) was used to detect the change of pretreated bagasse fibers. SEM images were taken of different pretreated bagasse samples with acceleration voltage of 2.0 KV. Results and discussion Proximate analysis Proximate analysis of SB sample (20-40 mesh) are presented in Table I. Primarily this analysis usually evaluate the fuel characteristics of raw materials. According to Sun et al., 2009, higher moisture and ashcontent in samples lessen the heating value. Ultimate analysis Ultimate analysis denotes the elemental configuration of SB such as carbon, hydrogen, oxygen, nitrogen and sulfur which are shown in Table II. This examination helps to measure the percentage of carbon and hydrogen content in biomass that is responsible to determine the amount of air is required for complete combustion, composition of combustion gases and heat is generated by it (Poddar et al. 2014). Chemical properties Raw SB contained 34.66±12% α-cellulose, 22.43±08% pentosan and 19.57±06% klason lignin in which 1.75±04% acid soluble lignin (dry basis) was detected by technical association of the pulp and paper industry (TAPPI) method. The chemical composition of SB was determined by acid hydrolysis and it was calculatedby HPLC method as 45.35% glucose and 30.64% xylose (Sujan et al., 2018). α-cellulose yield Bagasse is mainly composed of cellulose, hemicellulose and lignin. Besides these there are some extractives such as ash, wax, gum, pectin etc. During alkali pretreatment usually most of the extractives are removed with the increasing of alkali concentration. Pretreatment of SB with different alkali concentration based on raw material (0%, 2%, 4%, 6% and 8%), time (60, 90 and 120 min) and temperature (80, 100 and 120oC) are shown in Table III. Consequences of each independent experiment varying with alkali concentration, temperature and time on the α-cellulose yield were analyzed using MATLAB software. Apparently, it was observed that α-cellulose yield was considerably increased with changes of alkali concentration ranging from 0% to 8% (34.66% to 63.58%). But no noticeable variation was observed in temperature-time alteration during pretreatment of SB. Based on cellulose percentage obtained in treated bagasse, the optimum conditions for pretreatment reaction were selected as alkali concentration 8%, time 90 min and temperature 100oC. Although cellulose content in treated SB was slowly increased with the increase of alkali concentration 12% attime 90 min and temperature 100oC but as a consequence of huge chemical consumption, recovery problem, chance of losses cellulose and hemicellulose, alkali concentrations for pretreatment above 8% was not considered as ideal concentration. Effect of alkali concentration (AC), cooking time and temperature (temp) charge on α-cellulose yield as well as their statistical significance on the basis of F-test number are presented in regression equations (4). As shown in equations, cooking time at the maximum temperature had no significant effect on α-cellulose yield followed by alkali concentration charge. Effect of temperature on α-cellulose yield was less in employed cooking conditions. For α-cellulose yield: α-cellulose yield = 82.05-0.151×temp-1.56×time-0.69×AC (R2=0.89, adjusted R2=0.87) ............................................. (4) For Pentosan: Pentosan = 13.58+0.086 × tem p + 0.006 × time+0.445 × AC (R2=0.67, adjusted R2=0.63) ............................................. (5) For Lignin: Lignin = 27.662-0.041×temp-0.001×time-01.242×AC (R2=0.93, adjusted R2=0.92) ............................................. (6) For predicting α-cellulose, percentage of pentosan and lignin, the most influential factor was alkali concentration and then cooking temperature for α-cellulose yield, which exhibited an almost linear dependence on both operational variables. The coefficient of determinations is good for α-cel- lulose yield and lignin percentage which hovers around 90 percent, although the figure is moderate more than 60 percent for pentosan. All these three models are significant (p<0.05) at 5% level of significance. In order to perceive the impact of alkali concentration and temperature on these three parameters, three-dimensional (3D) response surface plots were created by plotting the response (α-cellulose yield) pentosan and lignin on the Z-axis versus the most influential one independent variable alkali concentration and temperature as shown in Fig. 3 (a), (b) and (c). XRD analysis In 19th century the cellulose crystalline structure has been discovered and later it was verified by X-ray crystallography (Meyer and Misch, 1937; Wilkie, 1961). The crystallinity index (CrI) of non-woody biomass, such as grasses and agricultural residues, varies depending on the type of biomass. Generally, non-woody biomass has a lower crystallinity index than woody biomass, typically ranging from 20-40%. The crystallinity index of non-woody biomass affects its digestibility and energy production potential.Recently, researchers are being paid more atten- tion on cellulose index because of its potential use in bioen- ergy production. Since then several different models of cellulose index have been proposed. The most popular two-phase cellulose model describes cellulose chains as containing both crystalline (ordered) and amorphous (less ordered) region (Park et al., 2010). Alkali treatment of bagasse has been found to increase the crystallinity index of the bagasse. Alkali treatment breaks down hemicellulose, making the cellulose molecules more ordered and crystalline, resulting in a higher crystallinity index. Alkali treatment also increases the digestibility of the bagasse, making it easier to break down and therefore more suitable for bioenergy production. X-ray diffraction spectra of the SB and alkali treated SB were presented in Figure 4. It was observed that the intensity of 101 and 002 peaks were gradually increased. The relative amount of crystalline cellu- lose (CrI) in the total solid were calculated based on the equation (2) and obtained 18.68%, 24.80%, 27.23%, 32.69%, 36.45% and 39.19% for raw SB, 0%, 2%, 4%, 6% and 8% alkali treated SB samples respectively. The CrI of alkali treated SB samples (not cellulose crystallinity) were intense- ly influenced by the composition of the samples. In case of lignocellulosic biomass examples, cellulose CrI measured the relative amount of crystalline cellulose in the total solid. Therefore, amorphous part of lignin and hemicellulose in biomass specimens were partially removed with the delig- nification process as a result the proportion of α-cellulose was increased and hence CrI would be increased gradually. This interpretation could be proved by the fact that alkali-treated SB had higher CrI than raw SB (Zhao et al. 2010) and the portion of cellulose in the treated SB was also increased gradually. According to the equation 3, the average sizes of crystallite obtained were 3.28, 3.51, 3.63, 3.78, 3.90 and 4.06 nm for raw SB, 0%, 2%, 4%, 6% and 8% alkali treated SB samples respectively. The experimental data revealed that during delignification, the size of crystallite was increased. SEM analysis SEM is one of the most commanding tools widely used to inves- tigate the lignocellulosic biomass surface (Amiri and Karimi, 2015). SEM is usually employed for surface characterization, morphology and inspection of microstructure. In respect of biomass example through SEM images we can compare the untreated and pretreated models which may lead to different insight into the biomass (Karimi and Taherzadeh, 2016). SEM provides two-dimensional images of raw SB and alkali treated SB, which were taken and compared to the outcome of NaOH treatment. All testers were coated with carbon tape and magnification of 500x was used. Figure 4 shows the wall of raw SB (Figure 6a) was intact where the alkali treated SB (Figures 6c-6f), the cell wall was ruptured or splitted and hence packing of the fibers were partially loosened (Firoz et al. 2012). Sugar and ethanol yield Enzymatic digestibility is the ability of enzymes to break down molecules into smaller molecules, such as glucose. It is used to measure the efficiency of enzyme-catalyzed reactions and the digestibility of cellulose. The enzymatic digestibility of a compound is affected by factors such as the compound's crystallinity index, the type of enzyme used, and the temperature and pH of the reaction. In this study the enzymatic digestibility of the alkali pretreated SB was improved by increasing pretreatment conditions. Typically the pretreatment settings are selected by considering various factors such as feedstock characteristics, pretreatment chemical cost, energy consumption and recovery efficiency (Wu et al. 2011). α-cellulose obtained by 8% alkali treated of SB was used for hydrolysis reaction and the reaction was carried out at an optimum condition set by Sujan et al. (2018). The most effective enzymatic hydrolysis was taken place with Trichoderma virideat 48 h and the theoretical yield of sugars i.e. glucose and xylose were obtained 89.59% and 61.23%, respectively. Through fermentation process generated sugars were used to check ethanol production. Yeast Saccharomyces cerevisiae presented worthy performance to convert C6 sugar into ethanol when it was incubated at 30°C for 24 h and 16.81±0.32% ethanol yield was detected. Conclusion α-cellulose yield was optimized by varying alkali concentration, temperature and cooking time. The most vital influencing factor for α-cellulose yield was alkali concentration and after that temperature performed a little bit. The optimum α-cellulose yield (63.58±0.05%) was obtained at 8% alkali concentration, 100oC and 90 minutes. In hydrolysis step of SB, Trichoderma viridaewas used to convert 8% alkali treated SB into sugars and it was attained 89.59% glucose and 61.23% xylose which gave higher yield in compare with our previous study such as ball milled and mesh size varied SB sample (Sujan et al. 2018). At fermentation step Saccha- romyces cerevisiae was used to convert hydrolysate of 8% alkali treated SB sample and ethanol yield was obtained 16.81±0.32% at 24 h. Compositional analysis, imaging and crystallinity are three methods were performed. SEM images can give different clues about SB including morphology, surface disruption and creation of highly accessible surface area iii) CrI (18.68% to 39.19%) and crystal size (3.28 nm to 4.06 nm) of SB are increased with different alkali treatment (0%-8%). In this case it was observed that crystallinity, crystal size, accessible surface area, porosity, particle size, lignin and hemicellulose content and enzyme adsorption/de- sorption were acted as the most impressive factors for digest- ibility of sugarcane bagasse. Acknowledgement Authors would like to acknowledge Pulp and Paper Division, BCSIR for their cooperation. Authors also like to thanks for the sincere assistance of Md. Abdul Hamid, Junior Techni- cian, IFRD, BCSIR. References Alvira P, Tomás-Pejó E, Ballesteros M and Negro MJ (2010), Pretreatment technologies for an efficient bioethanol production process based on enzymatic hydrolysis: a review, Bioresource technology 101(13): 4851-4861. DOI: org/10.1016/j.biortech.2009.11.093 Agbor VB, Cicek N, Sparling R, Berlin A and Levin, DB (2011), Biomass pretreatment: fundamentals toward application, Biotechnology advances 29(6): 675-685. DOI: org/10.1016/j.biotechadv.2011.05.005 Amiri H and Karimi K (2015), Improvement of acetone, butanol, and ethanol production from woody biomass using organosolv pretreatment, Bioprocess and biosys- tems engineering 38(10): 1959-1972. ASTM (2007), Standard test members for instrumental deter- mination of C, H, N in petroleum products and lubricants. ASTM D5291-02, ASTM international, West Conshohocken, PA. Canilha L, Chandel AK, Suzane dos Santos Milessi T, Antunes FAF, Luiz da Costa Freitas W, das Graças Almeida Felipe M and da Silva SS (2012), Bioconver- sion of sugarcane biomass into ethanol: an overview about composition, pretreatment methods, detoxifica- tion of hydrolysates, enzymatic saccharification and ethanol fermentation. Journal of Biomedicine and Biotechnology, DOI: org/10.1155/2012/989572 Den W, Sharma VK, Lee M, Nadadur G and Varma RS (2018), Lignocellulosic biomass transformations via greener oxidative pretreatment processes: access to energy and value-added chemicals, Frontiers in chem- istry 6: 141. DOI: org/10.3389/fchem.2018.00141 Ahmed FM, Rahman SR and Gomes DJ (2012), Saccharifi- cation of sugarcane bagasse by enzymatic treatment for bioethanol production, Malays J Microbiol 8(2), 97-103. Hirani AH, Javed N, Asif M, Basu SK and Kumar A (2018), A review on first-and second-generation biofuel productions. In Biofuels: greenhouse gas mitigation and global warming, Springer, New Delhi, pp 141-154. DOI: 10.1007/978-81-322-3763-1_8 Islam MZ, Asad MA, Hossain MT, Paul SC and Sujan SMA (2019), Bioethanol production from banana pseudostem by using separate and cocultures of cellu- lase enzyme with Saccharomyces cerevisiae, Journal of Environmental Science and Technology 12(4): 157-163. Jahan MS, Saeed A, Ni Y and He Z (2009), Pre-extraction and its impact on the alkaline pulping of bagasse, Journal of Biobased Materials and Bioenergy 3(4): 380-385. DOI: org/10.1166/jbmb.2009.1053 Janker-Obermeier I, Sieber V, Faulstich Mand Schieder D (2012), Solubilization of hemicellulose and lignin from wheat straw through microwave-assisted alkali treatment, Industrial Crops and Products 39: 198-203. DOI: org/10.1016/j.indcrop.2012.02.022 Karimi K and Taherzadeh MJ (2016), A critical review of analytical methods in pretreatment of lignocelluloses: composition, imaging, and crystallinity, Bioresource technology 200: 1008-1018. DOI: org/10.1016/j. biortech.2015.11.022 Maryana R, Ma’rifatun D, Wheni AI, Satriyo KW and Rizal WA (2014), Alkaline pretreatment on sugarcane bagasse for bioethanol production, Energy Procedia, 47: 250-254. DOI: org/10.1016/j.egypro.2014.01.221 McIntosh Sand Vancov T (2010), Enhanced enzyme sacchar- ification of Sorghum bicolor straw using dilute alkali pretreatment, Bioresource technology 101(17): 6718-6727. DOI: org/10.1016/j.biortech.2010.03.116 Meyer KH and Misch L (1937), Positions des atomes dans le nouveau modele spatial de la cellulose. Helvetica Chimica Acta 20(1): 232-244. DOI: org/10. 1002/hlca.19370200134 Park S, Baker JO, Himmel ME, Parilla PA and Johnson DK (2010), Cellulose crystallinity index: measurement techniques and their impact on interpreting cellulase performance, Biotechnology for biofuels 3(1): 10. Sujan SMA, Bari ML and Fakhruddin AN (2018), Effects of physical pretreatment (crushing and ball milling) on sugarcane bagasse for bioethanol production, Bangla- desh journal of botany 147(2): 257-64. Wilkie JS (1961), Carl Nägeli and the fine structure of living matter, Nature 190: 1145-1150. Wu L, Arakane M, Ike M, Wada M, Takai T, Gau Mand Tokuyasu K (2011), Low temperature alkali pretreat- ment for improving enzymatic digestibility of sweet sorghum bagasse for ethanol production, Bioresource Technology 102(7): 4793-4799. Zhang W, Okubayashi S and Bechtold T (2005), Fibrillation tendency of cellulosic fibers-Part 4. Effects of alkali pretreatment of various cellulosic fibers, Carbohydrate polymers. 61(4): 427-433. Zhao X, van der Heide E, Zhang Tand Liu D (2010), Delig- nification of sugarcane bagasse with alkali and peracetic acid and characterization of the pulp, BioRe- sources 5(3): 1565-1580. Zhao X, Zhang Land Liu D (2012), Biomass recalcitrance. Part II: Fundamentals of different pre‐treatments to increase the enzymatic digestibility of lignocellulose, Biofuels, Bioproducts and Biorefining 6(5): 561-579. Fig. 4. XRD spectra of a) 20-40 mesh size SB; b) 0% alkali treated sample; c) 2% alkali treated sample; d) 4% alkali treated sample; e) 6% alkali treated sample; f) 8% alkali treated sample 002 101 a b c d e f Fig. 5. Average size of crystalline for raw SB, 0, 2, 4, 6 and 8% alkali treated sample Optimization of alkali concentration in the pretreatment of sugarcane bagasse96 58(2) 2023 composed of vasculer bundles surrounded by non-fibrous parenchymatic cell (Jahan et al. 2009). Usually it is available as agricultural residue and byproduct from sugar mill indus- tries. In the year of 2015-16, there was 2,95,162 metric ton sugarcane bagasse was obtained from sugar mills of Bangla- desh (http://www.bsfic.gov.bd). To enrich cellulose content in biomass for enzymatic saccarification, current research has focused on various pretreatment processes. Currently, scientists have gone through a lot of studies on different pretreatment techniques to remove the compact and rigid composition by open up the cellulosic structure (Alvira et al. 2010). Pretreatment technologies are commonly abandoned to reduce the structural barriers and boost cellulose availabili- ty based on not only chemicals such as acid, alkali, oxidant, etc. but also several treatment settings. Alkali treatment is one of the most widespread and economical methods used for surface modification of lignocellulosic biomass. In the field of biorefinary, alkali pretreatment is intensively employed to develop the cellulosic materials both mechan- ically as well as chemically and such properties include tensile strength, dyeability, stability of dimension and reactivity (Wu et al. 2011). In alkali pretreatment biomass is treated under moderate reaction conditions ensuring inexpensiveness, inflated recycling possibilities of water and chemical agents (McIntosh and Vancov, 2010). Usually alkali treatment is the most effective pretreatment for agricultural residues, herbaceous crops and hardwood containing low lignin content in the compariosn of softwood containing high lignin content (Agbor et al. 2011; Canilha et al. 2012). This research aims is to investigate the alkaline pretreatment repercussions based on the chemical structure, surface morphology, structure and enzymatic digestion of sugarcane bagasse for sugars and ethanol production. Materials and methods Sugarcane bagasse (SB) was taken from street juice vendor. Warm tap water was used for washing purpose to eliminate the residual free sugars once again. This washing and pressing step was then repeated for three times. After that drying was done in an oven at 65oC for 16 h. Crushing of dried bagasse was done successively using a locally made crusher and sieved (Retsch, D-42759, HAAN, Germany) to have the particle size of 20-40 mesh on average. Before usage, the 20-40 mesh bagasse was kept at room temperature in an airtight plastic container (Sujan et al. 2018). Moisture and Ash content Moisture content of raw materials was measured using the ASTM D 4442-07 procedure. In a glass crucible, 2 grams of pretreatment sample containing 20-40 mesh were dried in oven at 105±2°C. The moisture content was expressed in percent wet basis and weight measurements were taken every 3 h. A muffle furnace was used for burning the Oven-dried samples at 575±25°C for ash content determination follow- ing the ASTM Standard E 1755-01. Volatile matter The volatile matter analysis was carried out in accordance with ASTM Standard D 271-48. Four grams of raw materials were heated in a furnace at 950±20°C for seven minutes. The weight loss, excluding the weight of moisture pushed off at 105°C, is then used to calculate volatile matter. Fixed carbon The difference between 100 and the total of volatile matter, moisture, and ash content was used to compute the fixed carbon percentage. Chemical analysis The technical association of the pulp and paper industry (TAPPI) method detects α-cellulose (T 203 cm-99), pentosan (T 223 cm-01), klason lignin (T211 om-83), and acid soluble lignin (T UM 250) on a dry basis. Ultimate analysis Ultimate analysis of samples was done by following the procedure ASTM Standard D 5291-02. Organic elemental analyzer (Flash 2000, Thermo Scientific, USA) was used with a specific condition (Reactor temp. 900°C, He: 250 kPa, O2: 250 kPa, TCD). Low concentration alkali pretreatment Different alkali concentration (0%, 2%, 4%, 6% and 8%), time (60, 90 and 120 min) and temperature (80, 100 and 120oC) were applied on raw material to check alkali effect on SB shown in Table 3. A portion of the ground bagasse 20-40 mesh (10g) was taken into plastic zipper bag (temperature 80oC, figure 2a and 2b) and stainless-steel reactor (tempera- ture 100oC and 120oC, figure 2c and 2d). Different alkali concentration was applied during the pretreatment process (alkali to SB ratio ranging between 1:12). The mixture was heated at a particular temperature such as in a water bath (80oC) and oil bath (100oC and 120oC) for a desired length of time. In the course of pretreatment process the sample was manually mixed 2-3 times to attain proper alkali treatment. The treated SB was placed into a polyester bag to remove excess alkali water by pressing it. After that it was vigorously washed with tap water (repeated for five times) to remove the remaining alkali. Finally the SB was dried in an oven at 65oC for 72 h and stored in a close container at room temperature for further experiment. Regression model Regression model has been developed for prediction of α-cel- lulose, pentosan and lignin in SB. The general form of the model is: y=α+α1x1 + α2x2+........................+αnxn + ε.......................(1) where α is the constant term. αi are the coefficient of variables xi. ε is the random error term which is minimized with Simple Least Squares Regression (SLSR). Regression coefficients of the independent variables namely alkali concentration, temperature and time are estimated by SLSR method for developing regression model to predict α-cellulose, pentosan and lignin. Efficiencies of these models are expressed by coefficient of multiple determination (R2) and Adjusted R2. 2.9 Separate hydrolysis and fermentation (SHF) The Hydrolysis experiment took place in 100 ml conical flux 10ml enzyme solution with 200 mg (2% dry wt.) in citrate buffer (0.05 M, pH 5.0) at 50°C for 48 h. In this case Trichoderma viride was used for hydrolysis. Hydrolysate was then heated for 15 min in a boiling water bath and centrifugation was done to remove solid particles. The supernatant was used for analysis of released sugars as described by Jamal et al. (2011). During fermentation process according to Firoz et al. (2012), 100 ml media was prepared and 0.5 g of commercial yeast Saccharomyces cerevisiae was used as inoculum which showed good performance to converts sugar into bioethanol. This inoculated media mixture was poured in a suitable glass- ware and was kept in a shaking incubator for 48 h. 10 ml of this medium was then added into the flask and it was properly covered with aluminum foil. Then it was placed in the incubator at 30oC for 24 h, 48 h, 72 h and 96 h for fermenta- tion of sugars to bio-ethanol according to Sujan et al. (2018). Samples from hydrolysis and fermentation were performed by HPLC. High performance liquid chromatography (HPLC) In characterize part, concentration of Sugars and ethanol were determined by HPLC (Ultimate 3000, Thermo Scien- tific, USA) method using Hyper Rez XP carbohydrate H+ 8 µm column (100×7.7 mm) equipped with a Refractive Index (Shodex RI-101) detector. The mobile phase was degassed with deionized water with a flow rate of 0.7 ml/min and column temperature was maintained at 70°C. It is possible to measure the total sugar concentration in the hydrolysis liquid fraction by comparing its peak area detected by HPLC with peak area of 1% standard sugar which consists of two sugars namely glucose and xylose (Sujan et al. 2018). The same column which is specialized for fermentation broth analysis is used for ethanol detec- tion. The kinetic parameters of ethanol fermentation were determined as follows (Islam et al. 2019): Crystallinity measurement X-ray diffraction (XRD) was used to determine the crystalline structure of the SB samples using a diffrac- tometer (GBC XRD) and filtered copper K radiation (λ = 0.1542 nm) by a monochromator at 35.50 kV voltage and 28 mA current, with a speed of about 2o/min and scan- ning in the range of 10 - 80ºC. The crystallinity index (CrI) was obtained from the ratio between the intensity of the 002 peak (I002, 2θ = 22.5) and the minimum dip (Iam, 2 θ = 18.5) according to the following equation (Roberta et al., 2012): CrI (%) = [(I002 - Iam)/I002] ×100 ...................................... (2) where I002 is the highest peak intensity of plane 002 and Iam is related to the amorphous structure. In present study, the average crystallite sizes were deter- mined from the Scherrer equation by using the diffraction pattern obtained from the 002 (hkl) lattice planes of cellulose samples D(hkl) = [(Kλ / B(hkl) cos2θ] ............................................... (3) Where D(hkl) (Crystallite size), K (Scherrer constant, 0.84), λ (X-ray wavelength, 0.154nm), B(hkl)(Full width half maxi- mum of the measured hkl reflection), and 2θ (Corresponding Bragg angle). Scanning electron microscopy (SEM) analysis In this research, SEM (ZEISS EVO 18 SEM) was used to detect the change of pretreated bagasse fibers. SEM images were taken of different pretreated bagasse samples with acceleration voltage of 2.0 KV. Results and discussion Proximate analysis Proximate analysis of SB sample (20-40 mesh) are presented in Table I. Primarily this analysis usually evaluate the fuel characteristics of raw materials. According to Sun et al., 2009, higher moisture and ashcontent in samples lessen the heating value. Ultimate analysis Ultimate analysis denotes the elemental configuration of SB such as carbon, hydrogen, oxygen, nitrogen and sulfur which are shown in Table II. This examination helps to measure the percentage of carbon and hydrogen content in biomass that is responsible to determine the amount of air is required for complete combustion, composition of combustion gases and heat is generated by it (Poddar et al. 2014). Chemical properties Raw SB contained 34.66±12% α-cellulose, 22.43±08% pentosan and 19.57±06% klason lignin in which 1.75±04% acid soluble lignin (dry basis) was detected by technical association of the pulp and paper industry (TAPPI) method. The chemical composition of SB was determined by acid hydrolysis and it was calculatedby HPLC method as 45.35% glucose and 30.64% xylose (Sujan et al., 2018). α-cellulose yield Bagasse is mainly composed of cellulose, hemicellulose and lignin. Besides these there are some extractives such as ash, wax, gum, pectin etc. During alkali pretreatment usually most of the extractives are removed with the increasing of alkali concentration. Pretreatment of SB with different alkali concentration based on raw material (0%, 2%, 4%, 6% and 8%), time (60, 90 and 120 min) and temperature (80, 100 and 120oC) are shown in Table III. Consequences of each independent experiment varying with alkali concentration, temperature and time on the α-cellulose yield were analyzed using MATLAB software. Apparently, it was observed that α-cellulose yield was considerably increased with changes of alkali concentration ranging from 0% to 8% (34.66% to 63.58%). But no noticeable variation was observed in temperature-time alteration during pretreatment of SB. Based on cellulose percentage obtained in treated bagasse, the optimum conditions for pretreatment reaction were selected as alkali concentration 8%, time 90 min and temperature 100oC. Although cellulose content in treated SB was slowly increased with the increase of alkali concentration 12% attime 90 min and temperature 100oC but as a consequence of huge chemical consumption, recovery problem, chance of losses cellulose and hemicellulose, alkali concentrations for pretreatment above 8% was not considered as ideal concentration. Effect of alkali concentration (AC), cooking time and temperature (temp) charge on α-cellulose yield as well as their statistical significance on the basis of F-test number are presented in regression equations (4). As shown in equations, cooking time at the maximum temperature had no significant effect on α-cellulose yield followed by alkali concentration charge. Effect of temperature on α-cellulose yield was less in employed cooking conditions. For α-cellulose yield: α-cellulose yield = 82.05-0.151×temp-1.56×time-0.69×AC (R2=0.89, adjusted R2=0.87) ............................................. (4) For Pentosan: Pentosan = 13.58+0.086 × tem p + 0.006 × time+0.445 × AC (R2=0.67, adjusted R2=0.63) ............................................. (5) For Lignin: Lignin = 27.662-0.041×temp-0.001×time-01.242×AC (R2=0.93, adjusted R2=0.92) ............................................. (6) For predicting α-cellulose, percentage of pentosan and lignin, the most influential factor was alkali concentration and then cooking temperature for α-cellulose yield, which exhibited an almost linear dependence on both operational variables. The coefficient of determinations is good for α-cel- lulose yield and lignin percentage which hovers around 90 percent, although the figure is moderate more than 60 percent for pentosan. All these three models are significant (p<0.05) at 5% level of significance. In order to perceive the impact of alkali concentration and temperature on these three parameters, three-dimensional (3D) response surface plots were created by plotting the response (α-cellulose yield) pentosan and lignin on the Z-axis versus the most influential one independent variable alkali concentration and temperature as shown in Fig. 3 (a), (b) and (c). XRD analysis In 19th century the cellulose crystalline structure has been discovered and later it was verified by X-ray crystallography (Meyer and Misch, 1937; Wilkie, 1961). The crystallinity index (CrI) of non-woody biomass, such as grasses and agricultural residues, varies depending on the type of biomass. Generally, non-woody biomass has a lower crystallinity index than woody biomass, typically ranging from 20-40%. The crystallinity index of non-woody biomass affects its digestibility and energy production potential.Recently, researchers are being paid more atten- tion on cellulose index because of its potential use in bioen- ergy production. Since then several different models of cellulose index have been proposed. The most popular two-phase cellulose model describes cellulose chains as containing both crystalline (ordered) and amorphous (less ordered) region (Park et al., 2010). Alkali treatment of bagasse has been found to increase the crystallinity index of the bagasse. Alkali treatment breaks down hemicellulose, making the cellulose molecules more ordered and crystalline, resulting in a higher crystallinity index. Alkali treatment also increases the digestibility of the bagasse, making it easier to break down and therefore more suitable for bioenergy production. X-ray diffraction spectra of the SB and alkali treated SB were presented in Figure 4. It was observed that the intensity of 101 and 002 peaks were gradually increased. The relative amount of crystalline cellu- lose (CrI) in the total solid were calculated based on the equation (2) and obtained 18.68%, 24.80%, 27.23%, 32.69%, 36.45% and 39.19% for raw SB, 0%, 2%, 4%, 6% and 8% alkali treated SB samples respectively. The CrI of alkali treated SB samples (not cellulose crystallinity) were intense- ly influenced by the composition of the samples. In case of lignocellulosic biomass examples, cellulose CrI measured the relative amount of crystalline cellulose in the total solid. Therefore, amorphous part of lignin and hemicellulose in biomass specimens were partially removed with the delig- nification process as a result the proportion of α-cellulose was increased and hence CrI would be increased gradually. This interpretation could be proved by the fact that alkali-treated SB had higher CrI than raw SB (Zhao et al. 2010) and the portion of cellulose in the treated SB was also increased gradually. According to the equation 3, the average sizes of crystallite obtained were 3.28, 3.51, 3.63, 3.78, 3.90 and 4.06 nm for raw SB, 0%, 2%, 4%, 6% and 8% alkali treated SB samples respectively. The experimental data revealed that during delignification, the size of crystallite was increased. SEM analysis SEM is one of the most commanding tools widely used to inves- tigate the lignocellulosic biomass surface (Amiri and Karimi, 2015). SEM is usually employed for surface characterization, morphology and inspection of microstructure. In respect of biomass example through SEM images we can compare the untreated and pretreated models which may lead to different insight into the biomass (Karimi and Taherzadeh, 2016). SEM provides two-dimensional images of raw SB and alkali treated SB, which were taken and compared to the outcome of NaOH treatment. All testers were coated with carbon tape and magnification of 500x was used. Figure 4 shows the wall of raw SB (Figure 6a) was intact where the alkali treated SB (Figures 6c-6f), the cell wall was ruptured or splitted and hence packing of the fibers were partially loosened (Firoz et al. 2012). Sugar and ethanol yield Enzymatic digestibility is the ability of enzymes to break down molecules into smaller molecules, such as glucose. It is used to measure the efficiency of enzyme-catalyzed reactions and the digestibility of cellulose. The enzymatic digestibility of a compound is affected by factors such as the compound's crystallinity index, the type of enzyme used, and the temperature and pH of the reaction. In this study the enzymatic digestibility of the alkali pretreated SB was improved by increasing pretreatment conditions. Typically the pretreatment settings are selected by considering various factors such as feedstock characteristics, pretreatment chemical cost, energy consumption and recovery efficiency (Wu et al. 2011). α-cellulose obtained by 8% alkali treated of SB was used for hydrolysis reaction and the reaction was carried out at an optimum condition set by Sujan et al. (2018). The most effective enzymatic hydrolysis was taken place with Trichoderma virideat 48 h and the theoretical yield of sugars i.e. glucose and xylose were obtained 89.59% and 61.23%, respectively. Through fermentation process generated sugars were used to check ethanol production. Yeast Saccharomyces cerevisiae presented worthy performance to convert C6 sugar into ethanol when it was incubated at 30°C for 24 h and 16.81±0.32% ethanol yield was detected. Conclusion α-cellulose yield was optimized by varying alkali concentration, temperature and cooking time. The most vital influencing factor for α-cellulose yield was alkali concentration and after that temperature performed a little bit. The optimum α-cellulose yield (63.58±0.05%) was obtained at 8% alkali concentration, 100oC and 90 minutes. In hydrolysis step of SB, Trichoderma viridaewas used to convert 8% alkali treated SB into sugars and it was attained 89.59% glucose and 61.23% xylose which gave higher yield in compare with our previous study such as ball milled and mesh size varied SB sample (Sujan et al. 2018). At fermentation step Saccha- romyces cerevisiae was used to convert hydrolysate of 8% alkali treated SB sample and ethanol yield was obtained 16.81±0.32% at 24 h. Compositional analysis, imaging and crystallinity are three methods were performed. SEM images can give different clues about SB including morphology, surface disruption and creation of highly accessible surface area iii) CrI (18.68% to 39.19%) and crystal size (3.28 nm to 4.06 nm) of SB are increased with different alkali treatment (0%-8%). In this case it was observed that crystallinity, crystal size, accessible surface area, porosity, particle size, lignin and hemicellulose content and enzyme adsorption/de- sorption were acted as the most impressive factors for digest- ibility of sugarcane bagasse. Acknowledgement Authors would like to acknowledge Pulp and Paper Division, BCSIR for their cooperation. Authors also like to thanks for the sincere assistance of Md. Abdul Hamid, Junior Techni- cian, IFRD, BCSIR. References Alvira P, Tomás-Pejó E, Ballesteros M and Negro MJ (2010), Pretreatment technologies for an efficient bioethanol production process based on enzymatic hydrolysis: a review, Bioresource technology 101(13): 4851-4861. DOI: org/10.1016/j.biortech.2009.11.093 Agbor VB, Cicek N, Sparling R, Berlin A and Levin, DB (2011), Biomass pretreatment: fundamentals toward application, Biotechnology advances 29(6): 675-685. DOI: org/10.1016/j.biotechadv.2011.05.005 Amiri H and Karimi K (2015), Improvement of acetone, butanol, and ethanol production from woody biomass using organosolv pretreatment, Bioprocess and biosys- tems engineering 38(10): 1959-1972. ASTM (2007), Standard test members for instrumental deter- mination of C, H, N in petroleum products and lubricants. ASTM D5291-02, ASTM international, West Conshohocken, PA. Canilha L, Chandel AK, Suzane dos Santos Milessi T, Antunes FAF, Luiz da Costa Freitas W, das Graças Almeida Felipe M and da Silva SS (2012), Bioconver- sion of sugarcane biomass into ethanol: an overview about composition, pretreatment methods, detoxifica- tion of hydrolysates, enzymatic saccharification and ethanol fermentation. Journal of Biomedicine and Biotechnology, DOI: org/10.1155/2012/989572 Den W, Sharma VK, Lee M, Nadadur G and Varma RS (2018), Lignocellulosic biomass transformations via greener oxidative pretreatment processes: access to energy and value-added chemicals, Frontiers in chem- istry 6: 141. DOI: org/10.3389/fchem.2018.00141 Ahmed FM, Rahman SR and Gomes DJ (2012), Saccharifi- cation of sugarcane bagasse by enzymatic treatment for bioethanol production, Malays J Microbiol 8(2), 97-103. Hirani AH, Javed N, Asif M, Basu SK and Kumar A (2018), A review on first-and second-generation biofuel productions. In Biofuels: greenhouse gas mitigation and global warming, Springer, New Delhi, pp 141-154. DOI: 10.1007/978-81-322-3763-1_8 Islam MZ, Asad MA, Hossain MT, Paul SC and Sujan SMA (2019), Bioethanol production from banana pseudostem by using separate and cocultures of cellu- lase enzyme with Saccharomyces cerevisiae, Journal of Environmental Science and Technology 12(4): 157-163. Jahan MS, Saeed A, Ni Y and He Z (2009), Pre-extraction and its impact on the alkaline pulping of bagasse, Journal of Biobased Materials and Bioenergy 3(4): 380-385. DOI: org/10.1166/jbmb.2009.1053 Janker-Obermeier I, Sieber V, Faulstich Mand Schieder D (2012), Solubilization of hemicellulose and lignin from wheat straw through microwave-assisted alkali treatment, Industrial Crops and Products 39: 198-203. DOI: org/10.1016/j.indcrop.2012.02.022 Karimi K and Taherzadeh MJ (2016), A critical review of analytical methods in pretreatment of lignocelluloses: composition, imaging, and crystallinity, Bioresource technology 200: 1008-1018. DOI: org/10.1016/j. biortech.2015.11.022 Maryana R, Ma’rifatun D, Wheni AI, Satriyo KW and Rizal WA (2014), Alkaline pretreatment on sugarcane bagasse for bioethanol production, Energy Procedia, 47: 250-254. DOI: org/10.1016/j.egypro.2014.01.221 McIntosh Sand Vancov T (2010), Enhanced enzyme sacchar- ification of Sorghum bicolor straw using dilute alkali pretreatment, Bioresource technology 101(17): 6718-6727. DOI: org/10.1016/j.biortech.2010.03.116 Meyer KH and Misch L (1937), Positions des atomes dans le nouveau modele spatial de la cellulose. Helvetica Chimica Acta 20(1): 232-244. DOI: org/10. 1002/hlca.19370200134 Park S, Baker JO, Himmel ME, Parilla PA and Johnson DK (2010), Cellulose crystallinity index: measurement techniques and their impact on interpreting cellulase performance, Biotechnology for biofuels 3(1): 10. Sujan SMA, Bari ML and Fakhruddin AN (2018), Effects of physical pretreatment (crushing and ball milling) on sugarcane bagasse for bioethanol production, Bangla- desh journal of botany 147(2): 257-64. Wilkie JS (1961), Carl Nägeli and the fine structure of living matter, Nature 190: 1145-1150. Wu L, Arakane M, Ike M, Wada M, Takai T, Gau Mand Tokuyasu K (2011), Low temperature alkali pretreat- ment for improving enzymatic digestibility of sweet sorghum bagasse for ethanol production, Bioresource Technology 102(7): 4793-4799. Zhang W, Okubayashi S and Bechtold T (2005), Fibrillation tendency of cellulosic fibers-Part 4. Effects of alkali pretreatment of various cellulosic fibers, Carbohydrate polymers. 61(4): 427-433. Zhao X, van der Heide E, Zhang Tand Liu D (2010), Delig- nification of sugarcane bagasse with alkali and peracetic acid and characterization of the pulp, BioRe- sources 5(3): 1565-1580. Zhao X, Zhang Land Liu D (2012), Biomass recalcitrance. Part II: Fundamentals of different pre‐treatments to increase the enzymatic digestibility of lignocellulose, Biofuels, Bioproducts and Biorefining 6(5): 561-579. Fig. 6(a-f). SEM images a) 20-40 mesh size SB; b) 0% alkali treated sample; c) 2% alkali treated sample; d) 4% alkali treated sample; e) 6% alkali treated sample; f) 8% alkali treated sample 6a 6b 6c 6d 6e 6f Ec: Ethanol concentration Table IV. The percentage of ethanol yield with different time of fermentation at 30oC Time (h) 24 h 48 h 72 h 96 h Ec (%) 16.81±0.32 13.25±0.16 12.12±0.20 10.41±0.23 composed of vasculer bundles surrounded by non-fibrous parenchymatic cell (Jahan et al. 2009). Usually it is available as agricultural residue and byproduct from sugar mill indus- tries. In the year of 2015-16, there was 2,95,162 metric ton sugarcane bagasse was obtained from sugar mills of Bangla- desh (http://www.bsfic.gov.bd). To enrich cellulose content in biomass for enzymatic saccarification, current research has focused on various pretreatment processes. Currently, scientists have gone through a lot of studies on different pretreatment techniques to remove the compact and rigid composition by open up the cellulosic structure (Alvira et al. 2010). Pretreatment technologies are commonly abandoned to reduce the structural barriers and boost cellulose availabili- ty based on not only chemicals such as acid, alkali, oxidant, etc. but also several treatment settings. Alkali treatment is one of the most widespread and economical methods used for surface modification of lignocellulosic biomass. In the field of biorefinary, alkali pretreatment is intensively employed to develop the cellulosic materials both mechan- ically as well as chemically and such properties include tensile strength, dyeability, stability of dimension and reactivity (Wu et al. 2011). In alkali pretreatment biomass is treated under moderate reaction conditions ensuring inexpensiveness, inflated recycling possibilities of water and chemical agents (McIntosh and Vancov, 2010). Usually alkali treatment is the most effective pretreatment for agricultural residues, herbaceous crops and hardwood containing low lignin content in the compariosn of softwood containing high lignin content (Agbor et al. 2011; Canilha et al. 2012). This research aims is to investigate the alkaline pretreatment repercussions based on the chemical structure, surface morphology, structure and enzymatic digestion of sugarcane bagasse for sugars and ethanol production. Materials and methods Sugarcane bagasse (SB) was taken from street juice vendor. Warm tap water was used for washing purpose to eliminate the residual free sugars once again. This washing and pressing step was then repeated for three times. After that drying was done in an oven at 65oC for 16 h. Crushing of dried bagasse was done successively using a locally made crusher and sieved (Retsch, D-42759, HAAN, Germany) to have the particle size of 20-40 mesh on average. Before usage, the 20-40 mesh bagasse was kept at room temperature in an airtight plastic container (Sujan et al. 2018). Moisture and Ash content Moisture content of raw materials was measured using the ASTM D 4442-07 procedure. In a glass crucible, 2 grams of pretreatment sample containing 20-40 mesh were dried in oven at 105±2°C. The moisture content was expressed in percent wet basis and weight measurements were taken every 3 h. A muffle furnace was used for burning the Oven-dried samples at 575±25°C for ash content determination follow- ing the ASTM Standard E 1755-01. Volatile matter The volatile matter analysis was carried out in accordance with ASTM Standard D 271-48. Four grams of raw materials were heated in a furnace at 950±20°C for seven minutes. The weight loss, excluding the weight of moisture pushed off at 105°C, is then used to calculate volatile matter. Fixed carbon The difference between 100 and the total of volatile matter, moisture, and ash content was used to compute the fixed carbon percentage. Chemical analysis The technical association of the pulp and paper industry (TAPPI) method detects α-cellulose (T 203 cm-99), pentosan (T 223 cm-01), klason lignin (T211 om-83), and acid soluble lignin (T UM 250) on a dry basis. Ultimate analysis Ultimate analysis of samples was done by following the procedure ASTM Standard D 5291-02. Organic elemental analyzer (Flash 2000, Thermo Scientific, USA) was used with a specific condition (Reactor temp. 900°C, He: 250 kPa, O2: 250 kPa, TCD). Low concentration alkali pretreatment Different alkali concentration (0%, 2%, 4%, 6% and 8%), time (60, 90 and 120 min) and temperature (80, 100 and 120oC) were applied on raw material to check alkali effect on SB shown in Table 3. A portion of the ground bagasse 20-40 mesh (10g) was taken into plastic zipper bag (temperature 80oC, figure 2a and 2b) and stainless-steel reactor (tempera- ture 100oC and 120oC, figure 2c and 2d). Different alkali concentration was applied during the pretreatment process (alkali to SB ratio ranging between 1:12). The mixture was heated at a particular temperature such as in a water bath (80oC) and oil bath (100oC and 120oC) for a desired length of time. In the course of pretreatment process the sample was manually mixed 2-3 times to attain proper alkali treatment. The treated SB was placed into a polyester bag to remove excess alkali water by pressing it. After that it was vigorously washed with tap water (repeated for five times) to remove the remaining alkali. Finally the SB was dried in an oven at 65oC for 72 h and stored in a close container at room temperature for further experiment. Regression model Regression model has been developed for prediction of α-cel- lulose, pentosan and lignin in SB. The general form of the model is: y=α+α1x1 + α2x2+........................+αnxn + ε.......................(1) where α is the constant term. αi are the coefficient of variables xi. ε is the random error term which is minimized with Simple Least Squares Regression (SLSR). Regression coefficients of the independent variables namely alkali concentration, temperature and time are estimated by SLSR method for developing regression model to predict α-cellulose, pentosan and lignin. Efficiencies of these models are expressed by coefficient of multiple determination (R2) and Adjusted R2. 2.9 Separate hydrolysis and fermentation (SHF) The Hydrolysis experiment took place in 100 ml conical flux 10ml enzyme solution with 200 mg (2% dry wt.) in citrate buffer (0.05 M, pH 5.0) at 50°C for 48 h. In this case Trichoderma viride was used for hydrolysis. Hydrolysate was then heated for 15 min in a boiling water bath and centrifugation was done to remove solid particles. The supernatant was used for analysis of released sugars as described by Jamal et al. (2011). During fermentation process according to Firoz et al. (2012), 100 ml media was prepared and 0.5 g of commercial yeast Saccharomyces cerevisiae was used as inoculum which showed good performance to converts sugar into bioethanol. This inoculated media mixture was poured in a suitable glass- ware and was kept in a shaking incubator for 48 h. 10 ml of this medium was then added into the flask and it was properly covered with aluminum foil. Then it was placed in the incubator at 30oC for 24 h, 48 h, 72 h and 96 h for fermenta- tion of sugars to bio-ethanol according to Sujan et al. (2018). Samples from hydrolysis and fermentation were performed by HPLC. High performance liquid chromatography (HPLC) In characterize part, concentration of Sugars and ethanol were determined by HPLC (Ultimate 3000, Thermo Scien- tific, USA) method using Hyper Rez XP carbohydrate H+ 8 µm column (100×7.7 mm) equipped with a Refractive Index (Shodex RI-101) detector. The mobile phase was degassed with deionized water with a flow rate of 0.7 ml/min and column temperature was maintained at 70°C. It is possible to measure the total sugar concentration in the hydrolysis liquid fraction by comparing its peak area detected by HPLC with peak area of 1% standard sugar which consists of two sugars namely glucose and xylose (Sujan et al. 2018). The same column which is specialized for fermentation broth analysis is used for ethanol detec- tion. The kinetic parameters of ethanol fermentation were determined as follows (Islam et al. 2019): Crystallinity measurement X-ray diffraction (XRD) was used to determine the crystalline structure of the SB samples using a diffrac- tometer (GBC XRD) and filtered copper K radiation (λ = 0.1542 nm) by a monochromator at 35.50 kV voltage and 28 mA current, with a speed of about 2o/min and scan- ning in the range of 10 - 80ºC. The crystallinity index (CrI) was obtained from the ratio between the intensity of the 002 peak (I002, 2θ = 22.5) and the minimum dip (Iam, 2 θ = 18.5) according to the following equation (Roberta et al., 2012): CrI (%) = [(I002 - Iam)/I002] ×100 ...................................... (2) where I002 is the highest peak intensity of plane 002 and Iam is related to the amorphous structure. In present study, the average crystallite sizes were deter- mined from the Scherrer equation by using the diffraction pattern obtained from the 002 (hkl) lattice planes of cellulose samples D(hkl) = [(Kλ / B(hkl) cos2θ] ............................................... (3) Where D(hkl) (Crystallite size), K (Scherrer constant, 0.84), λ (X-ray wavelength, 0.154nm), B(hkl)(Full width half maxi- mum of the measured hkl reflection), and 2θ (Corresponding Bragg angle). Scanning electron microscopy (SEM) analysis In this research, SEM (ZEISS EVO 18 SEM) was used to detect the change of pretreated bagasse fibers. SEM images were taken of different pretreated bagasse samples with acceleration voltage of 2.0 KV. Results and discussion Proximate analysis Proximate analysis of SB sample (20-40 mesh) are presented in Table I. Primarily this analysis usually evaluate the fuel characteristics of raw materials. According to Sun et al., 2009, higher moisture and ashcontent in samples lessen the heating value. Ultimate analysis Ultimate analysis denotes the elemental configuration of SB such as carbon, hydrogen, oxygen, nitrogen and sulfur which are shown in Table II. This examination helps to measure the percentage of carbon and hydrogen content in biomass that is responsible to determine the amount of air is required for complete combustion, composition of combustion gases and heat is generated by it (Poddar et al. 2014). Chemical properties Raw SB contained 34.66±12% α-cellulose, 22.43±08% pentosan and 19.57±06% klason lignin in which 1.75±04% acid soluble lignin (dry basis) was detected by technical association of the pulp and paper industry (TAPPI) method. The chemical composition of SB was determined by acid hydrolysis and it was calculatedby HPLC method as 45.35% glucose and 30.64% xylose (Sujan et al., 2018). α-cellulose yield Bagasse is mainly composed of cellulose, hemicellulose and lignin. Besides these there are some extractives such as ash, wax, gum, pectin etc. During alkali pretreatment usually most of the extractives are removed with the increasing of alkali concentration. Pretreatment of SB with different alkali concentration based on raw material (0%, 2%, 4%, 6% and 8%), time (60, 90 and 120 min) and temperature (80, 100 and 120oC) are shown in Table III. Consequences of each independent experiment varying with alkali concentration, temperature and time on the α-cellulose yield were analyzed using MATLAB software. Apparently, it was observed that α-cellulose yield was considerably increased with changes of alkali concentration ranging from 0% to 8% (34.66% to 63.58%). But no noticeable variation was observed in temperature-time alteration during pretreatment of SB. Based on cellulose percentage obtained in treated bagasse, the optimum conditions for pretreatment reaction were selected as alkali concentration 8%, time 90 min and temperature 100oC. Although cellulose content in treated SB was slowly increased with the increase of alkali concentration 12% attime 90 min and temperature 100oC but as a consequence of huge chemical consumption, recovery problem, chance of losses cellulose and hemicellulose, alkali concentrations for pretreatment above 8% was not considered as ideal concentration. Effect of alkali concentration (AC), cooking time and temperature (temp) charge on α-cellulose yield as well as their statistical significance on the basis of F-test number are presented in regression equations (4). As shown in equations, cooking time at the maximum temperature had no significant effect on α-cellulose yield followed by alkali concentration charge. Effect of temperature on α-cellulose yield was less in employed cooking conditions. For α-cellulose yield: α-cellulose yield = 82.05-0.151×temp-1.56×time-0.69×AC (R2=0.89, adjusted R2=0.87) ............................................. (4) For Pentosan: Pentosan = 13.58+0.086 × tem p + 0.006 × time+0.445 × AC (R2=0.67, adjusted R2=0.63) ............................................. (5) For Lignin: Lignin = 27.662-0.041×temp-0.001×time-01.242×AC (R2=0.93, adjusted R2=0.92) ............................................. (6) For predicting α-cellulose, percentage of pentosan and lignin, the most influential factor was alkali concentration and then cooking temperature for α-cellulose yield, which exhibited an almost linear dependence on both operational variables. The coefficient of determinations is good for α-cel- lulose yield and lignin percentage which hovers around 90 percent, although the figure is moderate more than 60 percent for pentosan. All these three models are significant (p<0.05) at 5% level of significance. In order to perceive the impact of alkali concentration and temperature on these three parameters, three-dimensional (3D) response surface plots were created by plotting the response (α-cellulose yield) pentosan and lignin on the Z-axis versus the most influential one independent variable alkali concentration and temperature as shown in Fig. 3 (a), (b) and (c). XRD analysis In 19th century the cellulose crystalline structure has been discovered and later it was verified by X-ray crystallography (Meyer and Misch, 1937; Wilkie, 1961). The crystallinity index (CrI) of non-woody biomass, such as grasses and agricultural residues, varies depending on the type of biomass. Generally, non-woody biomass has a lower crystallinity index than woody biomass, typically ranging from 20-40%. The crystallinity index of non-woody biomass affects its digestibility and energy production potential.Recently, researchers are being paid more atten- tion on cellulose index because of its potential use in bioen- ergy production. Since then several different models of cellulose index have been proposed. The most popular two-phase cellulose model describes cellulose chains as containing both crystalline (ordered) and amorphous (less ordered) region (Park et al., 2010). Alkali treatment of bagasse has been found to increase the crystallinity index of the bagasse. Alkali treatment breaks down hemicellulose, making the cellulose molecules more ordered and crystalline, resulting in a higher crystallinity index. Alkali treatment also increases the digestibility of the bagasse, making it easier to break down and therefore more suitable for bioenergy production. X-ray diffraction spectra of the SB and alkali treated SB were presented in Figure 4. It was observed that the intensity of 101 and 002 peaks were gradually increased. The relative amount of crystalline cellu- lose (CrI) in the total solid were calculated based on the equation (2) and obtained 18.68%, 24.80%, 27.23%, 32.69%, 36.45% and 39.19% for raw SB, 0%, 2%, 4%, 6% and 8% alkali treated SB samples respectively. The CrI of alkali treated SB samples (not cellulose crystallinity) were intense- ly influenced by the composition of the samples. In case of lignocellulosic biomass examples, cellulose CrI measured the relative amount of crystalline cellulose in the total solid. Therefore, amorphous part of lignin and hemicellulose in biomass specimens were partially removed with the delig- nification process as a result the proportion of α-cellulose was increased and hence CrI would be increased gradually. This interpretation could be proved by the fact that alkali-treated SB had higher CrI than raw SB (Zhao et al. 2010) and the portion of cellulose in the treated SB was also increased gradually. According to the equation 3, the average sizes of crystallite obtained were 3.28, 3.51, 3.63, 3.78, 3.90 and 4.06 nm for raw SB, 0%, 2%, 4%, 6% and 8% alkali treated SB samples respectively. The experimental data revealed that during delignification, the size of crystallite was increased. SEM analysis SEM is one of the most commanding tools widely used to inves- tigate the lignocellulosic biomass surface (Amiri and Karimi, 2015). SEM is usually employed for surface characterization, morphology and inspection of microstructure. In respect of biomass example through SEM images we can compare the untreated and pretreated models which may lead to different insight into the biomass (Karimi and Taherzadeh, 2016). SEM provides two-dimensional images of raw SB and alkali treated SB, which were taken and compared to the outcome of NaOH treatment. All testers were coated with carbon tape and magnification of 500x was used. Figure 4 shows the wall of raw SB (Figure 6a) was intact where the alkali treated SB (Figures 6c-6f), the cell wall was ruptured or splitted and hence packing of the fibers were partially loosened (Firoz et al. 2012). Sugar and ethanol yield Enzymatic digestibility is the ability of enzymes to break down molecules into smaller molecules, such as glucose. It is used to measure the efficiency of enzyme-catalyzed reactions and the digestibility of cellulose. The enzymatic digestibility of a compound is affected by factors such as the compound's crystallinity index, the type of enzyme used, and the temperature and pH of the reaction. In this study the enzymatic digestibility of the alkali pretreated SB was improved by increasing pretreatment conditions. Typically the pretreatment settings are selected by considering various factors such as feedstock characteristics, pretreatment chemical cost, energy consumption and recovery efficiency (Wu et al. 2011). α-cellulose obtained by 8% alkali treated of SB was used for hydrolysis reaction and the reaction was carried out at an optimum condition set by Sujan et al. (2018). The most effective enzymatic hydrolysis was taken place with Trichoderma virideat 48 h and the theoretical yield of sugars i.e. glucose and xylose were obtained 89.59% and 61.23%, respectively. Through fermentation process generated sugars were used to check ethanol production. Yeast Saccharomyces cerevisiae presented worthy performance to convert C6 sugar into ethanol when it was incubated at 30°C for 24 h and 16.81±0.32% ethanol yield was detected. Conclusion α-cellulose yield was optimized by varying alkali concentration, temperature and cooking time. The most vital influencing factor for α-cellulose yield was alkali concentration and after that temperature performed a little bit. The optimum α-cellulose yield (63.58±0.05%) was obtained at 8% alkali concentration, 100oC and 90 minutes. In hydrolysis step of SB, Trichoderma viridaewas used to convert 8% alkali treated SB into sugars and it was attained 89.59% glucose and 61.23% xylose which gave higher yield in compare with our previous study such as ball milled and mesh size varied SB sample (Sujan et al. 2018). At fermentation step Saccha- romyces cerevisiae was used to convert hydrolysate of 8% alkali treated SB sample and ethanol yield was obtained 16.81±0.32% at 24 h. Compositional analysis, imaging and crystallinity are three methods were performed. SEM images can give different clues about SB including morphology, surface disruption and creation of highly accessible surface area iii) CrI (18.68% to 39.19%) and crystal size (3.28 nm to 4.06 nm) of SB are increased with different alkali treatment (0%-8%). In this case it was observed that crystallinity, crystal size, accessible surface area, porosity, particle size, lignin and hemicellulose content and enzyme adsorption/de- sorption were acted as the most impressive factors for digest- ibility of sugarcane bagasse. Acknowledgement Authors would like to acknowledge Pulp and Paper Division, BCSIR for their cooperation. Authors also like to thanks for the sincere assistance of Md. Abdul Hamid, Junior Techni- cian, IFRD, BCSIR. References Alvira P, Tomás-Pejó E, Ballesteros M and Negro MJ (2010), Pretreatment technologies for an efficient bioethanol production process based on enzymatic hydrolysis: a review, Bioresource technology 101(13): 4851-4861. DOI: org/10.1016/j.biortech.2009.11.093 Agbor VB, Cicek N, Sparling R, Berlin A and Levin, DB (2011), Biomass pretreatment: fundamentals toward application, Biotechnology advances 29(6): 675-685. DOI: org/10.1016/j.biotechadv.2011.05.005 Amiri H and Karimi K (2015), Improvement of acetone, butanol, and ethanol production from woody biomass using organosolv pretreatment, Bioprocess and biosys- tems engineering 38(10): 1959-1972. ASTM (2007), Standard test members for instrumental deter- mination of C, H, N in petroleum products and lubricants. ASTM D5291-02, ASTM international, West Conshohocken, PA. Canilha L, Chandel AK, Suzane dos Santos Milessi T, Antunes FAF, Luiz da Costa Freitas W, das Graças Almeida Felipe M and da Silva SS (2012), Bioconver- sion of sugarcane biomass into ethanol: an overview about composition, pretreatment methods, detoxifica- tion of hydrolysates, enzymatic saccharification and ethanol fermentation. Journal of Biomedicine and Biotechnology, DOI: org/10.1155/2012/989572 Den W, Sharma VK, Lee M, Nadadur G and Varma RS (2018), Lignocellulosic biomass transformations via greener oxidative pretreatment processes: access to energy and value-added chemicals, Frontiers in chem- istry 6: 141. DOI: org/10.3389/fchem.2018.00141 Ahmed FM, Rahman SR and Gomes DJ (2012), Saccharifi- cation of sugarcane bagasse by enzymatic treatment for bioethanol production, Malays J Microbiol 8(2), 97-103. Hirani AH, Javed N, Asif M, Basu SK and Kumar A (2018), A review on first-and second-generation biofuel productions. In Biofuels: greenhouse gas mitigation and global warming, Springer, New Delhi, pp 141-154. DOI: 10.1007/978-81-322-3763-1_8 Islam MZ, Asad MA, Hossain MT, Paul SC and Sujan SMA (2019), Bioethanol production from banana pseudostem by using separate and cocultures of cellu- lase enzyme with Saccharomyces cerevisiae, Journal of Environmental Science and Technology 12(4): 157-163. Jahan MS, Saeed A, Ni Y and He Z (2009), Pre-extraction and its impact on the alkaline pulping of bagasse, Journal of Biobased Materials and Bioenergy 3(4): 380-385. DOI: org/10.1166/jbmb.2009.1053 Janker-Obermeier I, Sieber V, Faulstich Mand Schieder D (2012), Solubilization of hemicellulose and lignin from wheat straw through microwave-assisted alkali treatment, Industrial Crops and Products 39: 198-203. DOI: org/10.1016/j.indcrop.2012.02.022 Karimi K and Taherzadeh MJ (2016), A critical review of analytical methods in pretreatment of lignocelluloses: composition, imaging, and crystallinity, Bioresource technology 200: 1008-1018. DOI: org/10.1016/j. biortech.2015.11.022 Maryana R, Ma’rifatun D, Wheni AI, Satriyo KW and Rizal WA (2014), Alkaline pretreatment on sugarcane bagasse for bioethanol production, Energy Procedia, 47: 250-254. DOI: org/10.1016/j.egypro.2014.01.221 McIntosh Sand Vancov T (2010), Enhanced enzyme sacchar- ification of Sorghum bicolor straw using dilute alkali pretreatment, Bioresource technology 101(17): 6718-6727. DOI: org/10.1016/j.biortech.2010.03.116 Meyer KH and Misch L (1937), Positions des atomes dans le nouveau modele spatial de la cellulose. Helvetica Chimica Acta 20(1): 232-244. DOI: org/10. 1002/hlca.19370200134 Park S, Baker JO, Himmel ME, Parilla PA and Johnson DK (2010), Cellulose crystallinity index: measurement techniques and their impact on interpreting cellulase performance, Biotechnology for biofuels 3(1): 10. Sujan SMA, Bari ML and Fakhruddin AN (2018), Effects of physical pretreatment (crushing and ball milling) on sugarcane bagasse for bioethanol production, Bangla- desh journal of botany 147(2): 257-64. Wilkie JS (1961), Carl Nägeli and the fine structure of living matter, Nature 190: 1145-1150. Wu L, Arakane M, Ike M, Wada M, Takai T, Gau Mand Tokuyasu K (2011), Low temperature alkali pretreat- ment for improving enzymatic digestibility of sweet sorghum bagasse for ethanol production, Bioresource Technology 102(7): 4793-4799. Zhang W, Okubayashi S and Bechtold T (2005), Fibrillation tendency of cellulosic fibers-Part 4. Effects of alkali pretreatment of various cellulosic fibers, Carbohydrate polymers. 61(4): 427-433. Zhao X, van der Heide E, Zhang Tand Liu D (2010), Delig- nification of sugarcane bagasse with alkali and peracetic acid and characterization of the pulp, BioRe- sources 5(3): 1565-1580. Zhao X, Zhang Land Liu D (2012), Biomass recalcitrance. Part II: Fundamentals of different pre‐treatments to increase the enzymatic digestibility of lignocellulose, Biofuels, Bioproducts and Biorefining 6(5): 561-579. Sujan, Hossain, Uddin and Fakhruddin 97 composed of vasculer bundles surrounded by non-fibrous parenchymatic cell (Jahan et al. 2009). Usually it is available as agricultural residue and byproduct from sugar mill indus- tries. In the year of 2015-16, there was 2,95,162 metric ton sugarcane bagasse was obtained from sugar mills of Bangla- desh (http://www.bsfic.gov.bd). To enrich cellulose content in biomass for enzymatic saccarification, current research has focused on various pretreatment processes. Currently, scientists have gone through a lot of studies on different pretreatment techniques to remove the compact and rigid composition by open up the cellulosic structure (Alvira et al. 2010). Pretreatment technologies are commonly abandoned to reduce the structural barriers and boost cellulose availabili- ty based on not only chemicals such as acid, alkali, oxidant, etc. but also several treatment settings. Alkali treatment is one of the most widespread and economical methods used for surface modification of lignocellulosic biomass. In the field of biorefinary, alkali pretreatment is intensively employed to develop the cellulosic materials both mechan- ically as well as chemically and such properties include tensile strength, dyeability, stability of dimension and reactivity (Wu et al. 2011). In alkali pretreatment biomass is treated under moderate reaction conditions ensuring inexpensiveness, inflated recycling possibilities of water and chemical agents (McIntosh and Vancov, 2010). Usually alkali treatment is the most effective pretreatment for agricultural residues, herbaceous crops and hardwood containing low lignin content in the compariosn of softwood containing high lignin content (Agbor et al. 2011; Canilha et al. 2012). This research aims is to investigate the alkaline pretreatment repercussions based on the chemical structure, surface morphology, structure and enzymatic digestion of sugarcane bagasse for sugars and ethanol production. Materials and methods Sugarcane bagasse (SB) was taken from street juice vendor. Warm tap water was used for washing purpose to eliminate the residual free sugars once again. This washing and pressing step was then repeated for three times. After that drying was done in an oven at 65oC for 16 h. Crushing of dried bagasse was done successively using a locally made crusher and sieved (Retsch, D-42759, HAAN, Germany) to have the particle size of 20-40 mesh on average. Before usage, the 20-40 mesh bagasse was kept at room temperature in an airtight plastic container (Sujan et al. 2018). Moisture and Ash content Moisture content of raw materials was measured using the ASTM D 4442-07 procedure. In a glass crucible, 2 grams of pretreatment sample containing 20-40 mesh were dried in oven at 105±2°C. The moisture content was expressed in percent wet basis and weight measurements were taken every 3 h. A muffle furnace was used for burning the Oven-dried samples at 575±25°C for ash content determination follow- ing the ASTM Standard E 1755-01. Volatile matter The volatile matter analysis was carried out in accordance with ASTM Standard D 271-48. Four grams of raw materials were heated in a furnace at 950±20°C for seven minutes. The weight loss, excluding the weight of moisture pushed off at 105°C, is then used to calculate volatile matter. Fixed carbon The difference between 100 and the total of volatile matter, moisture, and ash content was used to compute the fixed carbon percentage. Chemical analysis The technical association of the pulp and paper industry (TAPPI) method detects α-cellulose (T 203 cm-99), pentosan (T 223 cm-01), klason lignin (T211 om-83), and acid soluble lignin (T UM 250) on a dry basis. Ultimate analysis Ultimate analysis of samples was done by following the procedure ASTM Standard D 5291-02. Organic elemental analyzer (Flash 2000, Thermo Scientific, USA) was used with a specific condition (Reactor temp. 900°C, He: 250 kPa, O2: 250 kPa, TCD). Low concentration alkali pretreatment Different alkali concentration (0%, 2%, 4%, 6% and 8%), time (60, 90 and 120 min) and temperature (80, 100 and 120oC) were applied on raw material to check alkali effect on SB shown in Table 3. A portion of the ground bagasse 20-40 mesh (10g) was taken into plastic zipper bag (temperature 80oC, figure 2a and 2b) and stainless-steel reactor (tempera- ture 100oC and 120oC, figure 2c and 2d). Different alkali concentration was applied during the pretreatment process (alkali to SB ratio ranging between 1:12). The mixture was heated at a particular temperature such as in a water bath (80oC) and oil bath (100oC and 120oC) for a desired length of time. In the course of pretreatment process the sample was manually mixed 2-3 times to attain proper alkali treatment. The treated SB was placed into a polyester bag to remove excess alkali water by pressing it. After that it was vigorously washed with tap water (repeated for five times) to remove the remaining alkali. Finally the SB was dried in an oven at 65oC for 72 h and stored in a close container at room temperature for further experiment. Regression model Regression model has been developed for prediction of α-cel- lulose, pentosan and lignin in SB. The general form of the model is: y=α+α1x1 + α2x2+........................+αnxn + ε.......................(1) where α is the constant term. αi are the coefficient of variables xi. ε is the random error term which is minimized with Simple Least Squares Regression (SLSR). Regression coefficients of the independent variables namely alkali concentration, temperature and time are estimated by SLSR method for developing regression model to predict α-cellulose, pentosan and lignin. Efficiencies of these models are expressed by coefficient of multiple determination (R2) and Adjusted R2. 2.9 Separate hydrolysis and fermentation (SHF) The Hydrolysis experiment took place in 100 ml conical flux 10ml enzyme solution with 200 mg (2% dry wt.) in citrate buffer (0.05 M, pH 5.0) at 50°C for 48 h. In this case Trichoderma viride was used for hydrolysis. Hydrolysate was then heated for 15 min in a boiling water bath and centrifugation was done to remove solid particles. The supernatant was used for analysis of released sugars as described by Jamal et al. (2011). During fermentation process according to Firoz et al. (2012), 100 ml media was prepared and 0.5 g of commercial yeast Saccharomyces cerevisiae was used as inoculum which showed good performance to converts sugar into bioethanol. This inoculated media mixture was poured in a suitable glass- ware and was kept in a shaking incubator for 48 h. 10 ml of this medium was then added into the flask and it was properly covered with aluminum foil. Then it was placed in the incubator at 30oC for 24 h, 48 h, 72 h and 96 h for fermenta- tion of sugars to bio-ethanol according to Sujan et al. (2018). Samples from hydrolysis and fermentation were performed by HPLC. High performance liquid chromatography (HPLC) In characterize part, concentration of Sugars and ethanol were determined by HPLC (Ultimate 3000, Thermo Scien- tific, USA) method using Hyper Rez XP carbohydrate H+ 8 µm column (100×7.7 mm) equipped with a Refractive Index (Shodex RI-101) detector. The mobile phase was degassed with deionized water with a flow rate of 0.7 ml/min and column temperature was maintained at 70°C. It is possible to measure the total sugar concentration in the hydrolysis liquid fraction by comparing its peak area detected by HPLC with peak area of 1% standard sugar which consists of two sugars namely glucose and xylose (Sujan et al. 2018). The same column which is specialized for fermentation broth analysis is used for ethanol detec- tion. The kinetic parameters of ethanol fermentation were determined as follows (Islam et al. 2019): Crystallinity measurement X-ray diffraction (XRD) was used to determine the crystalline structure of the SB samples using a diffrac- tometer (GBC XRD) and filtered copper K radiation (λ = 0.1542 nm) by a monochromator at 35.50 kV voltage and 28 mA current, with a speed of about 2o/min and scan- ning in the range of 10 - 80ºC. The crystallinity index (CrI) was obtained from the ratio between the intensity of the 002 peak (I002, 2θ = 22.5) and the minimum dip (Iam, 2 θ = 18.5) according to the following equation (Roberta et al., 2012): CrI (%) = [(I002 - Iam)/I002] ×100 ...................................... (2) where I002 is the highest peak intensity of plane 002 and Iam is related to the amorphous structure. In present study, the average crystallite sizes were deter- mined from the Scherrer equation by using the diffraction pattern obtained from the 002 (hkl) lattice planes of cellulose samples D(hkl) = [(Kλ / B(hkl) cos2θ] ............................................... (3) Where D(hkl) (Crystallite size), K (Scherrer constant, 0.84), λ (X-ray wavelength, 0.154nm), B(hkl)(Full width half maxi- mum of the measured hkl reflection), and 2θ (Corresponding Bragg angle). Scanning electron microscopy (SEM) analysis In this research, SEM (ZEISS EVO 18 SEM) was used to detect the change of pretreated bagasse fibers. SEM images were taken of different pretreated bagasse samples with acceleration voltage of 2.0 KV. Results and discussion Proximate analysis Proximate analysis of SB sample (20-40 mesh) are presented in Table I. Primarily this analysis usually evaluate the fuel characteristics of raw materials. According to Sun et al., 2009, higher moisture and ashcontent in samples lessen the heating value. Ultimate analysis Ultimate analysis denotes the elemental configuration of SB such as carbon, hydrogen, oxygen, nitrogen and sulfur which are shown in Table II. This examination helps to measure the percentage of carbon and hydrogen content in biomass that is responsible to determine the amount of air is required for complete combustion, composition of combustion gases and heat is generated by it (Poddar et al. 2014). Chemical properties Raw SB contained 34.66±12% α-cellulose, 22.43±08% pentosan and 19.57±06% klason lignin in which 1.75±04% acid soluble lignin (dry basis) was detected by technical association of the pulp and paper industry (TAPPI) method. The chemical composition of SB was determined by acid hydrolysis and it was calculatedby HPLC method as 45.35% glucose and 30.64% xylose (Sujan et al., 2018). α-cellulose yield Bagasse is mainly composed of cellulose, hemicellulose and lignin. Besides these there are some extractives such as ash, wax, gum, pectin etc. During alkali pretreatment usually most of the extractives are removed with the increasing of alkali concentration. Pretreatment of SB with different alkali concentration based on raw material (0%, 2%, 4%, 6% and 8%), time (60, 90 and 120 min) and temperature (80, 100 and 120oC) are shown in Table III. Consequences of each independent experiment varying with alkali concentration, temperature and time on the α-cellulose yield were analyzed using MATLAB software. Apparently, it was observed that α-cellulose yield was considerably increased with changes of alkali concentration ranging from 0% to 8% (34.66% to 63.58%). But no noticeable variation was observed in temperature-time alteration during pretreatment of SB. Based on cellulose percentage obtained in treated bagasse, the optimum conditions for pretreatment reaction were selected as alkali concentration 8%, time 90 min and temperature 100oC. Although cellulose content in treated SB was slowly increased with the increase of alkali concentration 12% attime 90 min and temperature 100oC but as a consequence of huge chemical consumption, recovery problem, chance of losses cellulose and hemicellulose, alkali concentrations for pretreatment above 8% was not considered as ideal concentration. Effect of alkali concentration (AC), cooking time and temperature (temp) charge on α-cellulose yield as well as their statistical significance on the basis of F-test number are presented in regression equations (4). As shown in equations, cooking time at the maximum temperature had no significant effect on α-cellulose yield followed by alkali concentration charge. Effect of temperature on α-cellulose yield was less in employed cooking conditions. For α-cellulose yield: α-cellulose yield = 82.05-0.151×temp-1.56×time-0.69×AC (R2=0.89, adjusted R2=0.87) ............................................. (4) For Pentosan: Pentosan = 13.58+0.086 × tem p + 0.006 × time+0.445 × AC (R2=0.67, adjusted R2=0.63) ............................................. (5) For Lignin: Lignin = 27.662-0.041×temp-0.001×time-01.242×AC (R2=0.93, adjusted R2=0.92) ............................................. (6) For predicting α-cellulose, percentage of pentosan and lignin, the most influential factor was alkali concentration and then cooking temperature for α-cellulose yield, which exhibited an almost linear dependence on both operational variables. The coefficient of determinations is good for α-cel- lulose yield and lignin percentage which hovers around 90 percent, although the figure is moderate more than 60 percent for pentosan. All these three models are significant (p<0.05) at 5% level of significance. In order to perceive the impact of alkali concentration and temperature on these three parameters, three-dimensional (3D) response surface plots were created by plotting the response (α-cellulose yield) pentosan and lignin on the Z-axis versus the most influential one independent variable alkali concentration and temperature as shown in Fig. 3 (a), (b) and (c). XRD analysis In 19th century the cellulose crystalline structure has been discovered and later it was verified by X-ray crystallography (Meyer and Misch, 1937; Wilkie, 1961). The crystallinity index (CrI) of non-woody biomass, such as grasses and agricultural residues, varies depending on the type of biomass. Generally, non-woody biomass has a lower crystallinity index than woody biomass, typically ranging from 20-40%. The crystallinity index of non-woody biomass affects its digestibility and energy production potential.Recently, researchers are being paid more atten- tion on cellulose index because of its potential use in bioen- ergy production. Since then several different models of cellulose index have been proposed. The most popular two-phase cellulose model describes cellulose chains as containing both crystalline (ordered) and amorphous (less ordered) region (Park et al., 2010). Alkali treatment of bagasse has been found to increase the crystallinity index of the bagasse. Alkali treatment breaks down hemicellulose, making the cellulose molecules more ordered and crystalline, resulting in a higher crystallinity index. Alkali treatment also increases the digestibility of the bagasse, making it easier to break down and therefore more suitable for bioenergy production. X-ray diffraction spectra of the SB and alkali treated SB were presented in Figure 4. It was observed that the intensity of 101 and 002 peaks were gradually increased. The relative amount of crystalline cellu- lose (CrI) in the total solid were calculated based on the equation (2) and obtained 18.68%, 24.80%, 27.23%, 32.69%, 36.45% and 39.19% for raw SB, 0%, 2%, 4%, 6% and 8% alkali treated SB samples respectively. The CrI of alkali treated SB samples (not cellulose crystallinity) were intense- ly influenced by the composition of the samples. In case of lignocellulosic biomass examples, cellulose CrI measured the relative amount of crystalline cellulose in the total solid. Therefore, amorphous part of lignin and hemicellulose in biomass specimens were partially removed with the delig- nification process as a result the proportion of α-cellulose was increased and hence CrI would be increased gradually. This interpretation could be proved by the fact that alkali-treated SB had higher CrI than raw SB (Zhao et al. 2010) and the portion of cellulose in the treated SB was also increased gradually. According to the equation 3, the average sizes of crystallite obtained were 3.28, 3.51, 3.63, 3.78, 3.90 and 4.06 nm for raw SB, 0%, 2%, 4%, 6% and 8% alkali treated SB samples respectively. The experimental data revealed that during delignification, the size of crystallite was increased. SEM analysis SEM is one of the most commanding tools widely used to inves- tigate the lignocellulosic biomass surface (Amiri and Karimi, 2015). SEM is usually employed for surface characterization, morphology and inspection of microstructure. In respect of biomass example through SEM images we can compare the untreated and pretreated models which may lead to different insight into the biomass (Karimi and Taherzadeh, 2016). SEM provides two-dimensional images of raw SB and alkali treated SB, which were taken and compared to the outcome of NaOH treatment. All testers were coated with carbon tape and magnification of 500x was used. Figure 4 shows the wall of raw SB (Figure 6a) was intact where the alkali treated SB (Figures 6c-6f), the cell wall was ruptured or splitted and hence packing of the fibers were partially loosened (Firoz et al. 2012). Sugar and ethanol yield Enzymatic digestibility is the ability of enzymes to break down molecules into smaller molecules, such as glucose. It is used to measure the efficiency of enzyme-catalyzed reactions and the digestibility of cellulose. The enzymatic digestibility of a compound is affected by factors such as the compound's crystallinity index, the type of enzyme used, and the temperature and pH of the reaction. In this study the enzymatic digestibility of the alkali pretreated SB was improved by increasing pretreatment conditions. Typically the pretreatment settings are selected by considering various factors such as feedstock characteristics, pretreatment chemical cost, energy consumption and recovery efficiency (Wu et al. 2011). α-cellulose obtained by 8% alkali treated of SB was used for hydrolysis reaction and the reaction was carried out at an optimum condition set by Sujan et al. (2018). The most effective enzymatic hydrolysis was taken place with Trichoderma virideat 48 h and the theoretical yield of sugars i.e. glucose and xylose were obtained 89.59% and 61.23%, respectively. Through fermentation process generated sugars were used to check ethanol production. Yeast Saccharomyces cerevisiae presented worthy performance to convert C6 sugar into ethanol when it was incubated at 30°C for 24 h and 16.81±0.32% ethanol yield was detected. Conclusion α-cellulose yield was optimized by varying alkali concentration, temperature and cooking time. The most vital influencing factor for α-cellulose yield was alkali concentration and after that temperature performed a little bit. The optimum α-cellulose yield (63.58±0.05%) was obtained at 8% alkali concentration, 100oC and 90 minutes. In hydrolysis step of SB, Trichoderma viridaewas used to convert 8% alkali treated SB into sugars and it was attained 89.59% glucose and 61.23% xylose which gave higher yield in compare with our previous study such as ball milled and mesh size varied SB sample (Sujan et al. 2018). At fermentation step Saccha- romyces cerevisiae was used to convert hydrolysate of 8% alkali treated SB sample and ethanol yield was obtained 16.81±0.32% at 24 h. Compositional analysis, imaging and crystallinity are three methods were performed. SEM images can give different clues about SB including morphology, surface disruption and creation of highly accessible surface area iii) CrI (18.68% to 39.19%) and crystal size (3.28 nm to 4.06 nm) of SB are increased with different alkali treatment (0%-8%). In this case it was observed that crystallinity, crystal size, accessible surface area, porosity, particle size, lignin and hemicellulose content and enzyme adsorption/de- sorption were acted as the most impressive factors for digest- ibility of sugarcane bagasse. Acknowledgement Authors would like to acknowledge Pulp and Paper Division, BCSIR for their cooperation. Authors also like to thanks for the sincere assistance of Md. Abdul Hamid, Junior Techni- cian, IFRD, BCSIR. References Alvira P, Tomás-Pejó E, Ballesteros M and Negro MJ (2010), Pretreatment technologies for an efficient bioethanol production process based on enzymatic hydrolysis: a review, Bioresource technology 101(13): 4851-4861. DOI: org/10.1016/j.biortech.2009.11.093 Agbor VB, Cicek N, Sparling R, Berlin A and Levin, DB (2011), Biomass pretreatment: fundamentals toward application, Biotechnology advances 29(6): 675-685. DOI: org/10.1016/j.biotechadv.2011.05.005 Amiri H and Karimi K (2015), Improvement of acetone, butanol, and ethanol production from woody biomass using organosolv pretreatment, Bioprocess and biosys- tems engineering 38(10): 1959-1972. ASTM (2007), Standard test members for instrumental deter- mination of C, H, N in petroleum products and lubricants. ASTM D5291-02, ASTM international, West Conshohocken, PA. Canilha L, Chandel AK, Suzane dos Santos Milessi T, Antunes FAF, Luiz da Costa Freitas W, das Graças Almeida Felipe M and da Silva SS (2012), Bioconver- sion of sugarcane biomass into ethanol: an overview about composition, pretreatment methods, detoxifica- tion of hydrolysates, enzymatic saccharification and ethanol fermentation. Journal of Biomedicine and Biotechnology, DOI: org/10.1155/2012/989572 Den W, Sharma VK, Lee M, Nadadur G and Varma RS (2018), Lignocellulosic biomass transformations via greener oxidative pretreatment processes: access to energy and value-added chemicals, Frontiers in chem- istry 6: 141. DOI: org/10.3389/fchem.2018.00141 Ahmed FM, Rahman SR and Gomes DJ (2012), Saccharifi- cation of sugarcane bagasse by enzymatic treatment for bioethanol production, Malays J Microbiol 8(2), 97-103. Hirani AH, Javed N, Asif M, Basu SK and Kumar A (2018), A review on first-and second-generation biofuel productions. In Biofuels: greenhouse gas mitigation and global warming, Springer, New Delhi, pp 141-154. DOI: 10.1007/978-81-322-3763-1_8 Islam MZ, Asad MA, Hossain MT, Paul SC and Sujan SMA (2019), Bioethanol production from banana pseudostem by using separate and cocultures of cellu- lase enzyme with Saccharomyces cerevisiae, Journal of Environmental Science and Technology 12(4): 157-163. Jahan MS, Saeed A, Ni Y and He Z (2009), Pre-extraction and its impact on the alkaline pulping of bagasse, Journal of Biobased Materials and Bioenergy 3(4): 380-385. DOI: org/10.1166/jbmb.2009.1053 Janker-Obermeier I, Sieber V, Faulstich Mand Schieder D (2012), Solubilization of hemicellulose and lignin from wheat straw through microwave-assisted alkali treatment, Industrial Crops and Products 39: 198-203. DOI: org/10.1016/j.indcrop.2012.02.022 Karimi K and Taherzadeh MJ (2016), A critical review of analytical methods in pretreatment of lignocelluloses: composition, imaging, and crystallinity, Bioresource technology 200: 1008-1018. DOI: org/10.1016/j. biortech.2015.11.022 Maryana R, Ma’rifatun D, Wheni AI, Satriyo KW and Rizal WA (2014), Alkaline pretreatment on sugarcane bagasse for bioethanol production, Energy Procedia, 47: 250-254. DOI: org/10.1016/j.egypro.2014.01.221 McIntosh Sand Vancov T (2010), Enhanced enzyme sacchar- ification of Sorghum bicolor straw using dilute alkali pretreatment, Bioresource technology 101(17): 6718-6727. DOI: org/10.1016/j.biortech.2010.03.116 Meyer KH and Misch L (1937), Positions des atomes dans le nouveau modele spatial de la cellulose. Helvetica Chimica Acta 20(1): 232-244. DOI: org/10. 1002/hlca.19370200134 Park S, Baker JO, Himmel ME, Parilla PA and Johnson DK (2010), Cellulose crystallinity index: measurement techniques and their impact on interpreting cellulase performance, Biotechnology for biofuels 3(1): 10. Sujan SMA, Bari ML and Fakhruddin AN (2018), Effects of physical pretreatment (crushing and ball milling) on sugarcane bagasse for bioethanol production, Bangla- desh journal of botany 147(2): 257-64. Wilkie JS (1961), Carl Nägeli and the fine structure of living matter, Nature 190: 1145-1150. Wu L, Arakane M, Ike M, Wada M, Takai T, Gau Mand Tokuyasu K (2011), Low temperature alkali pretreat- ment for improving enzymatic digestibility of sweet sorghum bagasse for ethanol production, Bioresource Technology 102(7): 4793-4799. Zhang W, Okubayashi S and Bechtold T (2005), Fibrillation tendency of cellulosic fibers-Part 4. Effects of alkali pretreatment of various cellulosic fibers, Carbohydrate polymers. 61(4): 427-433. Zhao X, van der Heide E, Zhang Tand Liu D (2010), Delig- nification of sugarcane bagasse with alkali and peracetic acid and characterization of the pulp, BioRe- sources 5(3): 1565-1580. Zhao X, Zhang Land Liu D (2012), Biomass recalcitrance. Part II: Fundamentals of different pre‐treatments to increase the enzymatic digestibility of lignocellulose, Biofuels, Bioproducts and Biorefining 6(5): 561-579. Optimization of alkali concentration in the pretreatment of sugarcane bagasse 58(2) 202398