Comparative study of the efficiency of pretreatment with alkaline hydrogen peroxide in pineapple bagasse in different granulometries submitted to acid and enzymatic saccharification

This work had the purpose of evaluating the efficiency of the pretreatment with alkaline hydrogen peroxide of pineapple bagasse in order to obtain fermentable sugars by applying acid and enzymatic hydrolysis to said residue. Four experimental designs were applied to study the best conditions for the pre-treatment. Total reducing sugars (TRS) concentration was the response and hydrogen peroxide concentration, time and temperature were the independent variables. The studies were conducted using pineapple bagasse with particle sizes of 20 mesh and 48 mesh. Acid saccharification, with 2.9% sulfuric acid (v/v), following the pre-treatment, yielded TRS concentrations that reached 0.094 g of TRS/g of raw bagasse for 20 mesh and 0.101 g of TRS/g of raw bagasse for 48 mesh. The enzymatic saccharification, with 9 FPU/g cellulase and 2% (m/v) of bagasse, reached 0.063 g of TRS/g of raw bagasse for both particle sizes. The peroxide concentration showed a significant influence, the use of high concentrations reduced the TRS output in both hydrolysis. With the results found in this work, it is possible to infer the feasibility of applying pineapple bagasse as a lignocellulosic raw material.


Introduction
There is a correlation between energy and prosperity, making energy indispensable for modern societies (Nakanishi et al., 2018). Renewable energy sources are an upcoming trend, and lignocellulosic biomass is recognized as a prominent raw material for the production of biofuels and other value-added products, thus it can be of value to the global energy supply (Chaturved & Verma, 2013). The conversion of biomass into various value-added biochemical products, including biofuel, is generally termed as biomass valorization (Ho et al., 2019). This phrase has been widely used in recent research publications, indicating the shift of global research focal point to the alternative sources for producing chemicals and bioenergy, which could be derived from the waste. Biofuels are already being used to compose part of the fuels in many countries. Although, their raw materials are, in most of the cases, also food sources. Second and third generation fuels come to give a noble end to lignocellulosic and lipid residues that, otherwise, would be increasing pollution of land and water sources. Climate crisis is something we can no longer ignore, we must find environmental friendly ways to produce energy right now (Aranda-Martinez et al., 2017;Silva et al., 2018;Santo et al., 2018).
Second generation ethanol presents an enormous growth potential. It does not depend on the increase of agricultural lands for its production and, it has the added advantage of not competing with the production of food, since it repurposes residues (Candido et al., 2019). Lignocellulosic biomass (LB) represents the main potential raw material for lignocellulosic feedstock biorefinery, due to its high polysaccharide content, high availability, and low cost (Travanini, 2016). These residues have cellulose and hemicellulose in its composition. These structures can be broken down in sugars, such as glucose and xylose, by means of saccharification (Boussarsar et al., 2009). However, the biomass conversion into biofuels is only economically viable if 100% of its carbon is efficiently utilized in the industrial process. Including the cellulose, constituted of glucose units, and hemicellulose, mostly pentose units, fractions (Bettiga et al., 2008).
According to Oliveira (2012), the main effect of exposing the lignocellulosic matrix to pretreatment agents for extended periods is cellulose loss, resulting in low output in the process. Although, this step is necessary to reduce cellulose's degree of crystallinity; dissociate the lignin-cellulose complex; increase biomass surface area; preserve pentoses maximizing the sugars output; and minimizing the generation of process inhibitors, in both saccharification and fermentation steps (Shen et al., 2011, Santos et al., 2018. This step is a major bottleneck in second-generation ethanol production. There is a plethora of methods to choose from, with specific effects on biomass components. So the method and conditions should be chosen according to the raw material and subsequent process steps (Maurya et al., 2015). The excellent delignification using alkaline peroxide is not the only reason why researchers are paying more attention in biomass valorization for the past decades. Unlike acid or alkaline pretreatment, alkaline peroxide pretreatment can be performed at relatively milder conditions (concentration, temperature) and at atmospheric pressure, while effectively removing lignin from various agricultural residues. Furthermore, alkaline peroxide is a relatively "green" reagent with low environmental impact as it can be easily decomposed to yield water and oxygen as end products. (Ho et al., 2019).
Alkaline peroxide pretreatment is more efficient than alkaline pretreatment to improve the lignin solubilization, and digestibility of the raw material .
Acid saccharification can occur with either concentrated or diluted acids. Sulfuric acid is the most commonly used one. However, other acids can be used, there are studies involving hydrochloric, nitric and phosphoric acids (Menon and Rao, 2012). Acid saccharification requires extreme temperatures and pressures. Opposed to that, enzymatic saccharification can occur at atmospheric pressures and temperatures around 50°C. It happens with the synergistic action among cellulolytic enzymes, and saves energy, besides being an expensive ingredient (Zhang et al., 2006). Enzymatic hydrolyses has the added advantage of not generating products that degrade the monomers, which will reduce the sugars output. Yet, these cellulolytic enzymes are unable to catalyze a complete and quick cellulose conversion. It means the conversion rates depend on the type of residues used, saccharification times, and enzymes concentration applied (Candido et al., 2019).
According to Dahunsi (2019), there is therefore need to evaluate the use of this alkaline pretreatment on other lignocelluloses as well to assess the optimal condition for the process. In 2017 Brazil was the third largest producer of pineapples in the world, behind Costa Rica and the Philippines (Shahbandeh, 2019). The consumption in natura represents 97% of all the production in Brazil. From the processed part, around 40% of the fruit is waste. The residues are solid and high in sucrose and fermenting sugars (Gil and Maupoey, 2018). They are also high in lignocellulosic material, with around 40% hemicellulose and 30% cellulose (Lousada Jr et al., 2006;Brito et al., 2020).
Based on the above, the objective of the present work was to study the performance of alkaline hydrogen peroxide pretreatment of the pineapple bagasse in two different granulometries. In order to do that the pretreatment was followed by diluted acid and enzymatic saccharification. The response analyzed was the total reducing sugars (TRS) output, to evaluate the application in producing second-generation (2G) ethanol.

Methodology
The present work was a quantitative experimental research. The authors Nogueira and Vasconcelos conducted the experimental work supervised by Castiglioni, Freitas, and Seolatto.

Biomass
Pineapple bagasse, composed of pulp and peel, was donated by the industry Doce Vida (Sweet Life), located in Anápolis -GO, Brazil. The material was dried in an air circulation oven at 60°C for 72 h. After drying, the material was milled and sieved. The fractions of 20 and 48 mesh were separated, and stored frozen (-6°C) in polyethylene containers.

Biomass Characterization
The pineapple bagasse, in its 20 and 48 mesh fractions, was characterized. The ash and moisture contents were tested according to the Instituto Adolfo Lutz (Adolfo Lutz Institute) (1976) methodology. The soluble and insoluble lignin contents were quantified using a methodology preconized by IUPAC as described by Oliveira (2012). The cellulose and hemicellulose contents were determined by methodology described by Browning (1967). The carbohydrate content was assessed by High Performance Liquid Chromatography (HPLC). The bagasse underwent an acid hydrolyses and the hydrolyzed portion was tested in a LC-20A Prominence Shimadzu chromatographer, with a SUPERCOGEL Ca column, coupled to a refraction detector. The carrier solvent was deionized water with a pump flow of 0.5 mL/min. The oven temperature was 80°C and the injection volume 20µL (OLIVEIRA, 2012). All the analysis were done in triplicates.

Pretreatment with alkaline hydrogen peroxide.
The pretreatment with alkaline hydrogen peroxide was conducted according to methodology described by Krishna (2000). The factorial design chosen to be applied in this work was a 2 3 central composite, which indicates a factorial experiment with three control factors (variables studied), each tested in two levels (encoded by -1 and 1). The experimental design matrix was generated using the software Statistica™. The selected factors were time (h), temperature (T), and concentration of hydrogen peroxide. The response variable was the total reducing sugars (TRS). To make the study more precise and analyze the repeatability it was added of 6 central points and 2 axial points, totalizing 17 assays. The variables with their encoded and actual values are presented on Table 1. The experimental design was run four times. The same conditions were studied for acid and enzymatic saccharifications for each granulometry. The assays were done in triplicates. Thus, 4 g of the homogenized dried biomass was put into 100 mL of hydrogen peroxide solution at different concentrations for the experimental design (Table 1). The pH was adjusted to 11.5 with sodium hydroxide (20 mol/L), since it is the pH that provides more precise results . The pretreatment reactions were conducted in 250 mL Erlenmyer flasks. Each flask was sealed hermetically before being put in an orbital shaker at 150 rpm. At the end of each reactional time, the samples were filtered and washed with distilled water until reaching neutral pH. After pre-treatment the material was subjected to acid and enzymatic hydrolysis to release reducing sugars which were quantified.

Acid and enzymatic saccharification
After filtration he solid part left from the pretreatment was dried at 40°C until constant weigh. In order to determine the best pretreatment conditions the acid and enzymatic saccharifications were conducted. At the end of each saccharification, the TRS were determined according to methodology described by Miller (1959) and the results were presented in g of TRS/g of raw bagasse. The results were analyzed using the software for windows Statistica 7.0™. The significance level considered for all the analyses was 10% (p-value<0.1).
For the acid saccharification, the material was transferred to 125 mL Erlenmeyer flasks. Then 25 mL of sulfuric acid (H2SO4) 2.9% m/v was added, and put in the autoclave for 30 minutes at 121°C (Moutta et al., 2011). After the autoclaving process the flasks were left to cool at room temperature. When they were cool, the solid portion was separated by vacuum filtration. The liquid fraction was neutralized with a sodium hydroxide solution (20 mol/L) before quantifying TRS.
The enzymatic saccharification assays applied Cellulase from Trichoderma reesei, obtained from Sigma Aldrich. The optimum pH and temperature were previously determined according to methodology described by Rabelo (2010). The dried pretreated bagasse was transferred to Erlenmeyer flasks. A buffer solution of sodium acetate pH 4.8 was added to each flask in enough volume to make a 2% (m/v) concentration of solids, considering the moisture content of the samples. The amount of cellulase was fixed at 9 FPU/g of dried biomass. The saccharification occurred at 50°C and 150 rpm in an orbital shaker. To determine the best saccharification time, aliquots were taken at times of 0, 3, 6, 12, 24, 48 and 60 h. The best reaction time was 48 h, the one used to calculate the results. The test also included a blank, not to overestimate the TRS results.

Biomass Characterization
The cellulose and hemicellulose polymers require cleaving into their monomers, which is done by hydrolysis (Kučerová and Výbohová, 2017). As we can see on Table 2, the pineapple bagasse presented reducing and non-reducing carbohydrates. Sucrose (non-reducing sugar) was the highest concentration monomer, for both fractions. Since sucrose, in certain conditions, like in acid hydrolyses, can breakdown into glucose and fructose, this can be an indication that the hydrolysis was not very drastic. Aziz et al., (2011), when characterizing pineapple juice found sucrose levels similar to the ones in this work Cassellis et al (2014), studied the structural, physical-chemical and functional properties of industrial pineapple residues to obtain information on the recommendation of its use as dietary fiber. Pineapple residues showed a high concentration of cellulose and low concentrations of lignin, which is an advantage when it is to be subjected to chemical or enzymatic hydrolysis.
The low amount of total of lignin, 23.03% and 22.57%, respectively to 20 and 48 mesh biomasses, permits the extraction of monomeric sugars from both the cellulosic and hemicellulosic portions (Kreangkrai Maneeintr et al., 2018).
Sugarcane, an attractive biomass for 2G ethanol for its lignin content, has similar lignin levels (20.28%) to the pineapple bagasse in this study (Da Silva et al., 2020). It is an indication that pineapple bagasse has potential to be efficient in this segment.
The results from both granulometries were statistically compared using Student's t-test, with a 90% confidence level.
The concentrations of glucose, xylose, arabinose, cellulose and hemicellulose in pineapple bagasse between the two granulometries were equal statistically. This shows that the damage degree does not affect the global composition of the raw material.
Silva (2011), obtained, in their study with pineapple bagasse, 31.69% for the cellulose fraction and 38.18% for hemicellulose applying the same methodology used in the present work. Maneeintr et al., (2018), had cellulose levels ranging from 25.23% to 27.58% in their work. The results in this work were higher, something that could be explained by the fact that the vegetable cells' composition vary depending on where they grow, rain levels, and other factors (Banerjee et al., 2018).

Evaluation of pretreatment for acid saccharification
For a profitable use of a potential source of sugars, cost effective pretreatment and hydrolysis have to be applied.
Usually the pretreatment is the most expensive step in the whole process (Bhattacharyya and Gundupalli, 2019). The results yielded by the design matrix applied for the pretreatment with alkaline hydrogen peroxide followed by acid hydrolysis are shown in Table 3.
As we can see, when keeping time and concentration in their lowest levels and the temperature at 50ºC the TRS output is practically unaltered for both granulometries. The best TRS response for the 20 mesh fraction occurred on assay 5, which had the lowest levels of temperature and concentration of H2O2 ([H2O2]). When comparing assay 5 with assay 1, shortest and longest time levels, the result is practically unchanged. This could suggest that reaction time, in the tested conditions, is not relevant for the process. Assays 9 and 10 also corroborate said indications. Assay 13, the one with lowest This is an indication that high concentrations of peroxide can be removing part of the cellulose and hemicellulose along with the lignin. Under normal circumstances, hydrogen peroxide can only react with the aliphatic part of lignin, while no changes and degradation of phenolic compounds are observed. However, hydrogen peroxide is able to attack the phenolic compounds when it is used under alkaline conditions and heated to a relatively high temperature, which will expose the phenolic ring and cause carboxylic groups to be added to the macromolecular structure. Such alteration of biomass structure by AHP pretreatment has led to effective lignin and xylan removals, while decreasing cellulose crystallinity as well as swelling of biomass due to the insertion of polar groups into the molecule (Ho et al., 2019).

Evaluation of the pretreatment of 20 mesh pineapple bagasse for acid saccharification.
The results were analyzed by multiple linear regression. The experimental results had a good data adjustment to the model (R²=97%) and the obtained model is presented in equation 1. Figure 1 presents the response surfaces generated with the proposed model.
[ART] = 0,067 -0,0055t 2 -0,0051T + 0,0024T 2 -0,017C + 0,003C*T [1] The surfaces show a higher TRS concentration in temperatures around 20ºC. This indicates that running the pretreatment close to room temperature would not jeopardize the reaction's output (assays 1 and 5). When analyzing the effect of time (Figure 1 A and B), we can see it did not have a significant interference in the TRS response. There are good results in a broad range of times (assays 1, 5 and 13). However, the hydrogen peroxide had the best results at the lowest concentrations (Figure 1 B and C). Neto et al., (2013) compared pretreatments with alkaline hydrogen peroxide and phosphoric acid followed by delignification with sodium hydroxide for sugarcane bagasse. The results were significantly higher for alkaline hydrogen peroxide. Source: Authors.

Evaluation of the pretreatment of 48 mesh pineapple bagasse for acid saccharification.
The multiple linear regression was also applied to this set of data and, after eliminating the non-significant parameters, the correlation coefficient (R²) was 99% and the obtained model is presented in equation 2.

Evaluation of pretreatment for enzymatic saccharification
The enzymatic saccharification process used cellulases for extracting sugar from the biomass. Cellulose and Hemicellulose in the biomass will be digested by this enzyme and hydrolyzed in to monosaccharides (Maneeintr et al., 2018).
The results obtained using the design matrix applied to perform the pretreatment with alkaline hydrogen peroxide followed by enzymatic saccharification with cellulases in both granulometries are shown in Table 4.  Table 4, we can see the highest output of TRS, for 20 mesh fraction, was in assay 1, and for 48 mesh in assay 13. Similar to the acid hydrolyses the best results were at the lowest conditions studied. Likewise, the lowest results occurred at the highest conditions of temperature (50°C), [H2O2] (6%), and time (24h).

Evaluation of the pretreatment of 20 mesh pineapple bagasse for enzymatic saccharification.
Many factors can influence the performance of enzymatic hydrolysis, namely the type of biomass, enzymatic load, the cocktail of enzymes, and hydrolysis time, but the severity of the pretreatment seems to be of the greatest importance (Pábon et al., 2020). In this work, the results were treated statistically and the non-significant parameters were eliminated from the model. After the adjustments the determination coefficient (R²) was 98% and the obtained model is presented in equation 3. Figure 3 presents the surfaces acquired from said model.
The evaluation of the graphs shows that the highest outputs came from the mildest studied conditions, in consonance with the data presented this far. Tests with identical conditions for time and H2O2, and temperature varying from level -1 to +1 (For example, between assays 1 and 3 or 2 and 4), presented similar results. It implicates a low interaction effect between these variable and temperature, as can be checked in equation 3.
When analyzing Figure 3, it appears that the best responses come from lower temperature uses. Observe, for example, assays 1 and 3, where all experiments were set to peroxide concentration 2%. Assay 1 (8h time and 20°C temperature) obtained 0.063 g TRS/g dry bagasse, while assay 3 (8 h time and 50°C temperature) had an output of 0.057 g TRS/g of dry bagasse, showing that the temperature had a greater negative effect, reducing the response value. This indicates that, for this pretreatment, lower temperature values are more favorable. It was also clear that the highest TRS outputs were obtained when the lowest levels of the variables were combined. This is corroborated by the results yielded by experiments 1, 2 and 5.
De Araújo et al., (2017), investigated the influence of chemical (Triton X-100) and biological surfactant preparation (rhamnolipids) in coconut husk hydrolysis that was subjected to pretreatment with alkaline hydrogen peroxide. They used as pre-treatment an alkaline hydrogen peroxide concentration close to the highest level used in this work (7.5%) followed by enzymatic hydrolysis and obtained 40% lignin content.

Evaluation of the pretreatment of 48 mesh pineapple bagasse for enzymatic saccharification.
From the results presented in Table 4, it was possible to analyze statistically the response behavior of each variable.
After the adjustments the determination coefficient (R²) was 97 % and the obtained model is presented in equation 4. Figure 4 presents the surfaces acquired from said model.
Enzymatic saccharification had similar results for both fractions analyzed. There was a constant positive effect on the application of low concentrations of hydrogen peroxide. It has also been observed in acid saccharification. Student's t test showed that most results were statistically equal, between fractions of 20 and 48 mesh. This indicates that there are no significant differences between the different particle sizes.
Acid saccharification, for both particle sizes, showed superior results when comparing Tables 2 and 3. Possibly, this is due to the use of only one enzyme, cellulase, in enzymatic saccharification. It is usual to suplement the hydrolysis with βglucosidase (Söderström et al., 2003), that has a high cost. According to Martin et al., (2007), there are some advantagens in applying acid hydrolyses in detriments of the enzymatic process. The chemical reagents cost less; there is a wide range of acids that can be applied; and its reactions are better understood, because there are more studies. Nevertheless, we have disadvantages as well, as the coproducts that can reduce the sugars output, and inhibit the fermentation. Some disadvantages of enzymatic saccharification are the need for large amounts of enzymes or long times, to increase the formation of glucose (Martins et al., 2015). Cao et al., (2016) evaluated the saccharification of sweet sorghum bagasse pretreated with hydrogen peroxide, applying Celullases and β Glucosidase and achieved a production of 40.26 g/L of TRS. This difference are likely due to the different biomasses or the combination of enzymes. Chen et al., (2009), state that alkaline pretreatment is a more efficient chemical process. When compared to acid pretreatments, it degrades sugars in a smaller degree. The authors compared alkaline hydrogen peroxide and sodium hydroxide. They observed that alkaline hydrogen peroxide is more effective in solubilizing lignin when compared to sodium hydroxide pretreatment. The lignin reduced from 23.3% to 5.6% with alkaline hydrogen peroxide and from 23.2% to 14.6% with sodium hydroxide. In the present work preliminary tests with sulfuric acid 2% for 24h at 20°C had a TRS yield of 0.032 g/g. It shows this method did not have satisfactory results when compared to the tests with alkaline hydrogen peroxide (0.094 g/g TRS, in assay 5).
According to Oliveira (2012), the main collateral effect from the exposure of lignocellulosic materials to hydrogen peroxide is cellulose loss. This will result in low process output. The results in this work show that alkaline hydrogen peroxide pretreatment is very effective. Although the best results occur at the lowest concentrations and shortest times. These is evidence that peroxide pretreatments may remove cellulose, as well as lignin. Therefore, further research is necessary to optimize the pretreatment conditions and maximize the sugars output.

Final Considerations
The pineapple bagasse characterization presents similar results to previous works on lignocellulosic residues. The total lignin 22.03% (20 mesh) and 22.57% (48 mesh) allows the extraction of monomeric sugars. Both types of saccharification had better responses at the lowest levels of alkaline hydrogen peroxide concentration, and time and temperatures did not have significant impacts. The fact that between the granulometries some of the experiments had statistically equal and for others statistically different outputs of reducing sugars impedes a conclusive analysis. It can indicate that the variations are more based on different portions of the biomass, than in the granulometry itself. Based on this data, the pineapple bagasse could be applied, not only as raw material for secondary processes, but as an alternative in producing 2G ethanol. Future studies on this subject should focus on locating the optimal contitions of pretreatment and hydrolysis to maximaze the glucose output.