Determination of Alpha-, Beta- and Gamma- Cellulose in Bagasse and Wheat Straw: Lignin Recovery, Characterization and Depolymerization


 Lignocellulosic biomass is an abundantly available byproduct obtained after the separation of edible parts from various crops that is a potential source to produce renewable and sustainable biofuels, chemicals, materials, and polymers, without altering the greenhouse gas emissions relative to fossil feedstocks. Valorisation of lignocellulosic biomass focuses on polysaccharides conversion to value-added chemicals and polymers. However, lignin rich of high carbon burned to generate energy and chemicals. For the development of an effective lignocellulosic biomass conversion technology to biofuels and chemicals, biomass composition analysis and their properties need to be characterized prior to biomass reactions, including polysaccharide hydrolysis and lignin depolymerization. In this work, we have determined alpha-, beta- and gamma- cellulose, pentosan, lignin, and silica percentages of wheat straw (WS) and two bagasse (BG I and II) samples. The impact of different types of biomass samples on composition, and lignin recovery by applying two-stage concentrated and dilute sulphuric acid treatment, has been discussed. Subsequent studies extended to the correlation of lignin properties and their susceptibility to depolymerization using homogeneous (1-methyl-3-(3-sulphopropyl)-imidazolium hydrogen sulphate) and heterogeneous (immobilized Brønsted acidic ionic liquid) catalysts to lower molar mass aromatic fractions. Thermal, physical, and chemical properties of WS, BG, and recovered lignin samples were characterized by using UV-visible, ATR, 13 C CP-MAS NMR, CHNS, XRD, and TGA techniques showed substantial differences in lignin structure and properties.


Introduction
Environmental concerns and depletion of ozone layers are motivated to develop an effective conversion technology that can produce renewable chemicals, fuels, and polymers from sustainable resources, including lignocellulosic biomass. Technologies currently used to produce biofuels are primarily based on edible biomass, including corn grains, sugar syrup, and starch. (Pimentel and Burgess, 2014;Tenenbaum, 2008). Edible biomass sources are contested with food supply, and their production are also limited to provide enough amounts of fuels, chemicals, and materials (Tenenbaum, 2008). Lignocellulose biomass, including rice husks, bagasse, corn stover, and wheat straw can be considered as a substantial source for the production of renewable chemicals, and fuels that can substantially help to reduce and/ or substitute the dependency on non-renewable fossil feedstocks (Lynd et al., 1991;Somerville, 2006).
Lignocellulosic biomass is primarily composed of three major biopolymers, including cellulose, hemicellulose, and lignin (Sandip K. Singh, 2020). Cellulose and hemicellulose are communally known as holocellulose (Ragauskas et al., 2006). Holocellulose is further classified to alpha-, beta-and gamma-cellulose (for more details see the T-203 cm-99 process) (Alpha-, beta-and gamma-cellulose in pulp, T 203 cm-99, 1999;Bray and Andrews, 1923). To determine the alpha-, beta-and gamma-cellulose, several processes were employed (e.g., cooking and bleaching). However, their values in lignocellulosic biomass varied with the types of substrate, temperature, and applied method (Wells, 1921). For the determination of alpha-, beta-and gamma-cellulose in the pulp, it is crucial to have the carbon balance during the biomass utilization regardless of chemicals or biofuels production. Additionally, the determination of cellulose values in plant biomass, substantially helps to analyse the degradation of polysaccharides and lignin during pulp and paper processing. Effective utilization of polysaccharides into biofuels, chemicals and functional carbon materials is known at both laboratory and industrial scales (Corma et al., 2007). However, utilization of lignin to valuable products is still a challenge due to its recalcitrant features and association with several linkages, including ether/ester (β-O-4, 5-O-4') and condensed C-C (β-5', β-β, 5-5') (Rinaldi et al., 2016;del Río et al., 2015;Mosier et al., 2005). Lignin is mainly generated as a byproduct from biorefinery, pulp and paper industries. It is commonly used as a low-grade material, including binder, additive, and burn to regenerate heat and chemicals used (Binder and Raines, 2010).
Global sugar production is roughly 166.18 million metric tons (MMT) in 2019/2020, and with an expected consumption of approximately 177.8 MMT for 2020/2021. India is the secondlargest sugar producer globally after Brazil in 2019/2020. India produces sugar roughly 28.9 MMT that is 16.3% worldwide (M. . The European Union produced the highest amounts of wheat (ca. 153.5 MMT) in 2019, and India produced approximately 102.19 MMT wheat, that quantity is the third largest wheat producer worldwide .
Both bagasse and wheat straw are the crop residues left after utilizing edible parts, i.e., sugar syrup and wheat grains from sugarcane and wheat, respectively. These residues are generally burned to generate heat and/ or used as cattle feed in India. However, the burning of renewable non-edible feedstocks is a concern to produce tons of toxic gases. To reduce and/ or eliminate the generation of toxic gases, and to find alternative solutions, developing an effective conversion technology of lignocellulosic biomass to chemicals, biofuels, and polymers are essential (Yanding Li et al., 2018;Alinejad et al., 2019).
Environmental and development are the main factors that affected the lignin structure over aging. Separation and isolation of lignin from biomass, are also considered that potentially altered the lignin structure during the processing (Sandip K. Singh and Dhepe, 2016b;2018b;Chaudhary and Dhepe, 2019;Vanholme et al., 2019). In our previous works (Sandip K. Singh and Dhepe, 2016b;2018b;Chaudhary and Dhepe, 2019) and others works (Constant et al., 2016;Sannigrahi et al., 2010), it was observed that lignin linkages, molar mass, functional groups and intensities varied either applying same procedure with different biomass or same biomass under different procedures (Sandip K. Singh, 2020).
To recover lignin from lignocellulosic biomass, several methods were reported using Kraft, soda, two-stage alkaline oxidative treatment, lignosulphate, dilute acid, enzymatic, ionic liquids, organosolv process and more (Schutyser et al., 2018;Sandip K. Singh et al., 2019;Bhalla et al., 2019;Sandip K. Singh, 2020). These processes operated at a range of reaction conditions, including temperature, time, chemical, organic solvents, ionic liquids, enzymes, pH, and alkali loadings (Mosier et al., 2005;Sandip K. Singh, 2020). The recovered lignin showed a wide range of variation regardless of physical, chemical, thermal or biological properties. These properties of lignin substantially influenced the conversion and utilization of lignin (Kozliak et al., 2016).
Bioethanol and biobutanol have been produced from concentrated sulphuric acid-treated lignocellulosic biomass, and the technologies used to generate biofuels at large scales, have been investigated by using concentrated sulphuric acid (Riaz et al., 2016). Lignin recovered by using concentrate sulphuric acid, associated with low contents of ether linkages, and has a high order of condensed linkages (carbon-carbon). The concentrate sulphuric acid hydrolysis lignin was depolymerized by using basic catalysts (e.g., sodium hydroxide, potassium hydroxide, and sodium carbonate) at high temperature (330 ℃), and process produced low yields of aromatic fractions due to high percentage of condensed linkages (Riaz et al., 2016). The concentrate sulphuric acid hydrolysis lignin was converted by using supercritical ethanol in presence of formic acid at 350 ℃ (Riaz et al., 2016). In our previous work, we recovered concentrated sulphuric acid hydrolysis lignin from coconut coir and characterized by using several analytical techniques, and recovered lignin was depolymerized by using a solid base catalyst (NaX) at 200 ℃ (Chaudhary and Dhepe, 2019). Ionic liquid (IL) is considered as Green solvents or catalysts due to their specific properties, including low vapor pressure, non-corrosive relative to mineral acids or bases, good thermal stability, and more properties (Sandip K. Singh and Savoy, 2020). Considering the specific properties of ILs, they have been applied in various reactions, including lignocellulosic biomass deconstruction, lignin depolymerization to aromatic monomers and more (Sandip K. Singh and Dhepe, 2016a;Bhalla et al., 2019).
In summary, previous studies suggested that several types of acid delignification methods can be used to recover lignin from lignocellulosic biomass. In this work, we performed two-stage concentrated and diluted sulfuric acid polysaccharides hydrolysis to recover lignin from two types of bagasse (BG I and II) and wheat straw (WS) crop residues. We extended our study to determine the composition, including alpha-, beta-and gamma-cellulose, pentosan, inorganic nutrients, and lignin of BG (I and II) and WS samples. A wide range of bulk and molecular levels analytical techniques including thermal gravimetric analysis (TGA), inductively coupled plasma -optical emission spectrometry (ICP-OES), scanning electron microscopy attached with energy dispersive X-ray spectroscopy (SEM-EDX), attenuated total reflection (ATR), ultra violet-visible spectroscopy (UV-Vis), X-ray diffraction (XRD), and 13 C cross polarized magic angle spinning ( 13 C CP-MAS) NMR were applied to correlate the impact of crop residues over the lignin recovery and properties. In our previous works, we have screen a set of different ILs, and the optimized catalysts used for this work (Sandip K. Singh  and used without any further purification unless mentioned.

Lignin recovery
Recovery of concentrated sulfuric acid insoluble lignin from bagasse (BG I and II) and wheat straw (WS), was done using a well-known method with a few minor modifications (Sannigrahi et al., 2008;Samuel et al., 2010). Oven dried sample (~1.0 g) was taken in a beaker (100 mL capacity) and added H2SO4 (15 mL, 72 % wt/wt) solution. The sample was retained in a water bath at 30±1 ℃ for 1 h. The reaction mixture was stirred with a glass rod with a pause of 10 min. Thereafter completion of reaction, mixture was diluted with hot distilled water (80 o C) and poured this sample in a 1000 mL round bottom flask (RB). The volume of reaction mixture was maintained up-to 600 mL by adding additional hot water. The reaction mixture was refluxed in an oil bath for 4 h, it was then transferred in a water bath at 50 C, for 16 h. The precipitate was filtered by using a G2 crucible and washed with hot water to remove the contamination of sulphuric acid and impurities. The filtrate was processed for UV-Vis spectroscopy to quantify the acid soluble lignin (for more details, see Section S1.5, ESM). The recovered concentrated sulfuric acid hydrolysis lignin as a precipitate was air dried for 6 h and transferred then in an oven at 60±2 ℃, 6 h. Finally, precipitate was transferred in a vacuum oven at 100±2 ºC, below 1.01 bar, for 4 h to remove moisture. The recovered precipitate is named a concentrated sulphuric acid hydrolysis lignin. The recovered lignin sample was thoroughly characterized by using a series of analytical techniques.

Quantification of alpha-, beta-and gamma-cellulose
Polysaccharides mainly composed of alpha-cellulose, beta-cellulose, and gamma-cellulose, and hemi/cellulose are known as holocellulose in layered plant cell walls. The compositional analysis including cellulose (alpha-, beta-and gamma-cellulose), pentosan, lignin, ash, and silica, of BGs and WS samples, was done by using a known method (T-203 cm-99 process) (Alpha-, beta-and gamma-cellulose in pulp, T 203 cm-99, 1999;Bray and Andrews, 1923),

Characterization
We collected BG and WS crop residues, and thoroughly characterized in our previous studies (Sandip K. Singh and Dhepe, 2018b;2016b). In this work, we have used several bulk and molecular levels techniques, including elemental analysis, TGA, ICP-OES, SEM-EDX, ATR, UV-Vis, XRD, and 13 C CP-MAS NMR, to understand the physical, chemical and thermal properties of crop residues and recovered lignin samples (for more details, sample preparation and instruments, see Section S2, ESM). The properties of recovered lignin, were correlated and investigated for conversion to low molar mass fractions by using homogeneous

Composition analysis
Plant biomass is mainly composed of cellulose, hemicellulose, and lignin in maximum amounts, whereas, silica, extractive, and inorganic nutrients are present in minimum amounts.
The utilization of cellulose and hemicellulose for biofuels, chemicals, polymers and more, is a well procured process at the industrial level (Sandford and Baird, 1983). The major advantages of using bagasse (BG) and wheat straw (WS), are their profuse availability in India. These crop residues are eventually burned to generate heat, and this process generates tons of toxic gases.
In this work, two BGs and WS samples were screened for acidic lignin recovery, characterization and further depolymerization to low molar mass of phenolic products. Figure   1 shows the composition of these biomass samples. BG samples show the higher amounts of quantified cell wall structural biopolymers (i.e., holocellulose (alpha-, beta-and gammacellulose) and lignin (89±1%)), and contain low amounts of silica. Based on the approximate composition of two BG samples, it is assumed that the structural properties of BGs (e.g., cell wall association, mass density, lignin structures, etc.) do not statistically differ. It could be then hypothesized that these BG samples can exhibit similar behaviour in terms of characterization and susceptibility for conversion. (i.e., alpha-, beta-and gamma-cellulose) fraction, and (c) silica content in ash fraction.

Lignin and crop residues
To determine how the type of biomass, including similar and different species impact over the recovered lignin properties, and ultimately, their suitability for conversion to low molar mass   (Tang et al., 2012) for crystalline, and ~16, ~18 o , (Wu et al., 2010;Oh et al., 2005) for amorphous, were characterized. A wide range of (12.5-32.5 o ) XRD patterns characterized in recovered lignin, that confirmed the recovered lignin samples are amorphous (Sarkanen, 1963). A sharp peak in WS derived lignin relative to BGs, observed at 26.59 o corresponds to Si (JCPDS file No. 33-1161). WS residue contained maximum amounts of ash contents ( Fig. 1), that included maximum silica (Fig. 1c). The presence of maximum amounts of silica, that is in good agreement with an XRD intense peak at 26.59 o . As observed from SEM analysis is a unique and versatile technique to define the structure of crop residues and lignin (Ghaffar and Fan, 2013). SEM images of BGs and WS samples were taken and showed uni-structural morphologies (images not shown). Spherical morphologies were seen for lignin derived from bagasse, whereas lignin derived from WS, showed a non-structural image. The structural difference of recovered lignin from different crop samples, can plausible explain due to the presence of high amounts of inorganic contents (Fig. 1) Singh and Dhepe, 2016b; 2018b). Finally, the weight loss from ~350 °C to ~600 °C observed due to the decomposition of aromatic moieties in the lignin structure. Thereafter, no further weight loss was observed. Approximately 2.5% solid remained as a constant weight until 800 °C. The remaining weight was ash content present in the lignin sample. The weight of left residues is also in agreement with the weight of ash, observed with composition analysis of WS sample. In addition, the presence of inorganic content and metal amounts in crop residues, is confirmed and quantified by ICP-OES analysis (Table S2, ESM) respectively.
The elemental mapping, including carbon, hydrogen, and oxygen is required for characterizing the chemical contents in biomass. The energy released during the combustion or oxidation process is directly correlated to the sum of carbon and hydrogen contents as a function of energy values in biomass. In contrast, the maximum amounts of oxygen and nitrogen contents have low heat values, and that can decrease the energy efficacy of materials. The presence of chromophoric groups in lignin, is characterized using UV-Vis absorption spectroscopy. Lignin has a high UV absorption due to the presence of various functional groups and aromatic moieties. These groups and moieties have a high rate of Π vacant orbitals. The recovered lignin samples were processed for UV-Vis absorption study, and the obtained results are shown in Fig. 3. Lignin is associated with high amounts of aromatic units, and these units have several Π vacant orbitals. Therefore, the more intense peak is observed around 205 nm that associates to Π-Π* transition of aromatic, alkene, or alkyne unit (Antosiewicz and Shugar, 2016). The appearance of mono/di-substituted aromatic phenolic rings (e.g., hydroxyl, methoxy, aryl oxide groups, etc.) in lignin, is observed around 230 nm. Free and/ or etherified hydroxyl groups attached to the phenolic structure are observed around 280 nm (Musha and Goring, 1975). The absorption band ~318 nm indicates the presence of aromatic conjugated structure (C-α, C=C and C=O) in recovered lignin samples (Shulga et al., 2012). The presence of different types of functional groups in recovered lignin was analysed using an ATR (Alfa Bruker) technique. Fig. 4 shows the ATR spectra of recovered lignin from BGs and WS crop residues with a typical band assignment from 750-1850 and 2750-3800 cm -1 . A comparison study of lignin was done based on the literature (Long et al., 2012;Weiying Li et al., 2011;Strassberger et al., 2015;Hergert, 1971). A band at 3364 cm -1 was assigned to the hydroxyl groups attached to phenolic, or side chain aliphatic unit. The peaks at 2923 (intense peak in BG I) and 2852 сm -1 were characterized for C-H stretching vibrations in the methoxy, methyl, or methylene group. The band ~1733 cm -1 assigned to the presence of unconjugated, conjugated carbonyl or aromatic carboxyl groups with almost similar intensities in all lignin samples. The bands at 1642, 1602, 1508, 1454 and ~1420 cm -1 , were assigned for stretching of benzene ring in syringyl, guaiacyl and p-coumaryl alcohol moieties. The presence of 1212 cm -1 and 1043 cm -1 peaks was noted for guaiacyl units, (Hergert, 1971) and these peaks are more intense in lignin derived from BG I. The peak at 1172 cm -1 assigned for p-coumaryl alcohol.
The appearance of these peaks indicates the recovered lignin samples were generally associated with G and H moieties. In addition, absorbance peaks of G unit (1212 and 1043 cm -1 ) in lignin samples, were characterized more intense to H (1167 cm -1 ). These results showed that the guaiacyl moiety is present in high concentrations in recovered lignin (Hussin et al., 2013). The appearance of more intense peaks ~ 280 nm in UV-Vis, also affirmed that guaiacyl moiety is present with high concentration relative to S and H moieties (Fig. 3). A shoulder peak was noted with almost similar intensities at ~1359 cm -1 in lignin recovered from BG II and WS samples. A strong band at 1087 cm -1 characterized for C-H deformation of secondary alcohols, or by C=O stretching vibrations (For more details of peak assignments and respective wavenumber see Table 2).  The solution state or solid-state NMR study was used to characterize the lignin morphology (Hatfield et al., 1987). Recovered lignin samples were subjected to solid-state 13 C CP-MAS NMR using 46.7 µs acquisition time (AQ), and 21,000 scans. A Bruker Avance-300 MHz spectrometers operating at 75.47 MHz frequency, was used. Approximately 200 mg dried lignin was taken to record 13 C CP-MAS NMR spectra (Section S2, ESM). An adamantine molecule as an internal standard was used for 13 C CP-MAS NMR. A comparison study of 13 C CP-MAS NMR for recovered lignin samples is done with available literature (Nimz et al., 1981;Holtman et al., 2006;Almendros et al., 1992).
Alkyl side chains including -CH3, -CH2-, -CH-, -C-, CH3CO-, etc., are appeared in maximum numbers in recovered lignin-that is derived from BG II crop residue whereas the most intense peaks are presented in WS derived lignin (For more details of 13 C CP-MAS spectra and peak assignments, see Fig. 5 and Table 3).
Methoxy (-OCH3) groups were noted in all lignin samples with variable peak intensities. BG I and WS recovered lignin appeared almost similar intensities whereas low intense peak for methoxy groups, observed for BG II derived lignin. 13 C NMR peak for Aα, Aβ or Aγ in β-O-4' substructures, presented in recovered lignin samples with variable intensities. The Cβ carbon in the dibenzodioxocin substructure appeared in lignin recovered from WS. 15 The peak for C8 (T8) and C'2,6 carbons in tricin (T'2, 6) substructure, appeared in all lignin. Fig.   5 shows the more intense signal for tricin (T'2, 6) Table 3 for more detail).

Lignin depolymerization
Several methods at lab and pilot scales have been investigated to valorise the polysaccharides into various chemicals through a biorefinery process (Wettstein et al., 2012;Diwan et al., 2020). Lignin was generated as a byproduct through a biorefinery process and is generally burned to generate both heat and used chemicals. Lignin valorisation is a substantial biorefinery pathway to reach the goal of maximum amounts of carbon efficacy.
We have investigated several methods including solid acids (Deepa and Dhepe, 2015), solid bases, (Chaudhary and Dhepe, 2017) Fig. 6 shows the maximum amounts of organic solvent-soluble product yield obtained with WS-derived lignin. Un-structural morphologies (SEM images, Fig. S1, ESM), and the appearance of a high intense peak ~1100 cm -1 for alkyl aryl ethers (C-O-C) (Rashid et al., 2016;Larkin, 2011) (Fig. 4), could be a plausible explanation for maximum yields from WS derived lignin. In a mechanistic study, the rate of acid-catalysed cleavage of ether bonds particularly β-O-4, occurred second order of magnitude faster for hydrolysis relative to carbon-carbon bonds (Sturgeon et al., 2014). Additionally, lignin-derived from BG (I and II) has shown almost similar organic solvents soluble product yields.
We have screened a series of reaction conditions to have maximum amounts of low molecular weight aromatic products using an I-BAIL catalyst. Under the optimized reaction conditions, the autogenerated pressure (32 bar) due to methanol and water (5:1 v/v) reaction medium, was noted (Sandip K. Singh and Dhepe, 2018c). Lignin-derived from WS showed (Fig. 6) the maximum amounts of organic solvent soluble products yield. It could be plausible explained based on the morphologies and a high intense ATR band at ~1100 cm -1 for alky aryl ether linkages (Fig. 4). That indicates higher the ether linkages in lignin structure yielded the maximum amounts of lignin depolymerization product (Sturgeon et al., 2014). Moreover, lignin recovered from BG (I and II) crop residues showed similar catalytic results of organic solvent soluble products.

Conclusions
In summary, quantification of alpha-, beta-and gamma-cellulose in crop residues, is a substantial way to balance the carbon efficacy during the production of biofuels or chemicals from polysaccharides. Two-stage concentrated, and dilute sulfuric acid hydrolysis lignin was recovered, characterized and correlated the impacts of their properties for depolymerization.
These two studies in this work can add more economical values to the biorefinery process for biofuels or chemical production. Our experimental designs and significant findings in this work, are shown as below; 1. Three crop residues (two BG I and II and WS) were processed for determination of alpha-, beta-and gamma-cellulose, pentosan, nutrients, silica, and lignin.
5. 13 C CP-MAS NMR showed the appearance of tricin in all lignin samples while the dibenzodioxocin and spirodienone substructures are appeared in WS derived lignin.
6. Lignin recovered from WS showed the maximum amounts of depolymerized organic solvent soluble products yield by using both homogenous and heterogeneous catalysts.
7. The maximum amounts of organic soluble products yield correlated with the lignin morphologies (SEM) and the presence of ether and/or ester linkage in lignin (ATR).