Effect of Chemical Treatment Sequence on Pineapple Peel Fiber: Chemical Composition, Thermal Stability and Thermal Degradation Kinetics

Paulo Henrique Fernandes Pereira (  fernandes_eng@yahoo.com.br ) UNESP FEG: Universidade Estadual Paulista Julio de Mesquita Filho Faculdade de Engenharia Campus de Guaratingueta https://orcid.org/0000-0002-9741-2161 Heitor Luiz Ornaghi Universidade Federal da Integracao Latino-Americana Daniel Magalhães de Oliveira UNESP FEG: Universidade Estadual Paulista Julio de Mesquita Filho Faculdade de Engenharia Campus de Guaratingueta Barbara Pereira USP EEL: Universidade de Sao Paulo Escola de Engenheria de Lorena Valdeir Arantes USP EEL: Universidade de Sao Paulo Escola de Engenheria de Lorena Maria Odila Hilário Cio UNESP FEG: Universidade Estadual Paulista Julio de Mesquita Filho Faculdade de Engenharia Campus de Guaratingueta


Introduction
Increasing attention has been devoted to sustainability for moving towards eco-friendly products (Sabarinathan et al. 2020). A agroindustrial residues represent one of the most promising alternatives due to their remarkable characteristics such as low cost, biodegradability, abundance, ease of handling, and availability, besides being a renewable resource (Vinod et al. 2020;Haldar and Purkait 2020;Debnath et al. 2021). In addition, agricultural residues like fruits and vegetables have been attracting attention from researchers and industries due to their potential to be converted into higher value bioproduct and to enhance the social and environmental bene ts (Wang and Zhao 2021). However, about 40-50% of waste of root crops, fruits, and vegetables produced worldwide is discarded (Salehi and Aghajanzadeh 2020).
According to Oculi et al. 2020, pineapple (Ananas comosus (L) Merr.) is ranked the third most important tropical fruit crop worldwide after banana and citrus. As recently reported by the Food and Agriculture Organization (FAO), about 26 million pineapple tons were produced globally in 2019 (Hadidi et al. 2020).
Brazil is the third pineapple producer worldwide, with plantations distributed throughout the territory and an estimated production of 2650.48 metric tons, behind the Philippines with 2671.71 metric tons and Costa Rica with 3056.45 metric tons produced in 2018 (Statista 2020). In the industry, pineapples are used for the manufacture of juices, jam, pulp, and syrup (Wu and Shiau 2015;Braga et al. 2020), with 20% of the whole pineapple fruit used for juice (Aruna 2019). However, about 35% of the total fresh pineapple weight is discarded (Dai and Huang 2017) after fruit processing (Kumar et al. 2021). A small portion of this residue may be used as livestock feed and fertilizer (Ketnawa et al. 2012), but most of it is burned as waste, resulting in environmental such as air pollution from smoke and haze (Gnanasekaran et al. 2021).Thus, pineapple peel is an abundantly available potential source of cellulose that has been investigated for applications such as bioethanol production (Casabar et al. 2020), hydrogel (Dai et al. 2020), plasticizer (Rodsamran and Sothornvit 2019), adsorption (Shakya and Agarwal 2019), nanocellulose (Dai et al. 2019) and plasticizer (Rodsamran and Sothornvit 2019).
Many studies have been reported in the literature regarding pretreatments of agricultural residues for different purposes (Marques et al. 2020;Putrino et al. 2020;Contreras-Zarazúa et al. 2021;Yu et al. 2021). Speci cally, alkaline hydrolysis is usually made using diluted or concentrated NaOH solution at low or high temperature (Bandyopadhyay-Ghosh et al., 2015, Cavali et al., 2020Fonseca et al., 2019). In the presence of NaOH, the intramolecular ester bonds between lignin and hemicellulose are saponi ed, resulting in the extraction of lignin (Kim et al. 2016). Potassium hydroxide (KOH) is also used in alkali pretreatment for different feedstocks (Xie et al. 2018) to remove the remaining hemicellulose content (Bhagia et al. 2018). After the pretreatment, a second stage is necessary to remove various chromophore groups (from lignin) and impurities present in the cellulose bers. There are different types of pretreatment found in the literature (Hongrattanavichit and Aht-Ong 2020); Flores-Velázquez et al., 2020;Vasco-Correa et al., 2019;Liu et al., 2019) with the primary goal of reducing the biomass in size and open its physical structure. In recent years, there is increasing use of H 2 O 2 as a total or partial substitute for chlorine-based bleaching agents with elemental chlorine-free and totally chlorine-free sequences because of the compound's adverse environmental effects (Khristova et al. 2002).
Few studies have reported the effect of the pretreatment on the thermal kinetic behavior after cellulose separation using an eco-friendly treatment chemical. Therefore, the aim of this study is to understand the in uence of pretreatments: alkaline treatment (NaOH 4% m/m) followed by a bleaching stage with H 2 O 2 /NaOH, from pineapple peel on the physicochemical properties of its products separated as cellulose and to provide basic data for the utilization of biomass. To study the kinetic parameters, such as activation energy (Ea), thermogravimetric analysis (TGA) is used. The materials were characterized by scanning electron microscopy (SEM), chemical composition, thermogravimetric analysis (TGA), X-ray diffraction (XRD), Fourier transformed infrared spectroscopy (ATR-FTIR), and other means to provide a theoretical basis for the mechanism of kinetic studied 2. Material And Methods

Reagents
The chemicals used in this work were sodium hydroxide 97% w/w (NaOH), hydrogen peroxide 30% w/w (H 2 O 2 ) purchased from Dinâmica, and potassium hydroxide 97% w/w (KOH) purchased from Vetec. The reagents were used as received without further puri cation.

Material
Pineapple peel ber (PPF) from Ananas comosus L. Merryl pineapple used in this work was ripe fruit residue collected at a street market at Guaratinguetá city (São Paulo State, Brazil). Fresh PPF was cleaned with tap water, cut into small pieces, and oven-dried at 60 °C for 48 h to remove all the moisture.
Then, the PPFs were chopped into sizes of 2-4 cm in a cutter (GP 1500 AB) and milled to a powder (size less than 35 mesh) using a primary knife grinder (WILLYE TE-650).

Chemical treatments
2.3.1 Alkaline treatment (AT) Brie y, 15 g of dried PPF was suspended with 900 mL of an aqueous NaOH 4% (w/v) at 70 °C for 1 h in ratio to solvent of 1:20, at 500 rpm in 2000 mL beaker. After 1 h, the treated bers were ltered using a Buchner Funnel with a 28 μm pore lter, and the solid fraction was washed with distilled water until pH 7 and oven-dried at 60 ºC for 24 h to obtain the alkaline treated (AT) bers.

First bleaching with H
10g ried AT ber was added in 600 mL of an aqueous NaOH solution (4% w/v) and 60 mL of H 2 O 2 (30% v/v) at 70 ºC, under mechanical stirring at 4000 rpm. After 1 h of reaction, 100 mL of NaOH (4% w/v) and 60 mL of H 2 O 2 (30% v/v) were added into the reaction mixture. The same procedure was repeated each hour, up to 3 h. At the end of the 3 h, the bers were vacuum ltered using a 28 μm pore lter and waterrinsed, washed with distilled water until pH ~7, and oven-dried at 60 ºC for 24 h.

Second bleaching H
Fibers from the 1B treatment were stirred in a solution of 600 mL NaOH (4% w/v) and 60 mL of H 2 O 2 (30% v/v) at 70 ºC, under mechanical stirring at 4000 rpm. The same procedure was repeated as described in item 2.3.2, except that the reaction was carried out for 2 h of reaction, hereafter referred to as 2B.

Chemical composition
All samples were determined in triplicate. Fresh pineapple peel, AT, 1B, and 2B A were chemically analyzed following the NREL protocol (National Renewable Energy Laboratory) (Sluiter et al. 2012) to determine the percentages of lignin (soluble and insoluble), hemicellulose, and extractives in the bers before and after each chemical treatment. Monomeric sugars were quanti ed by HPLC (Waters using an HPX-87H column (Bio-Rad)) at 45 °C and eluted with 5 mmol/L H 2 SO 4 at 0.6 mL/min. Acid-soluble lignin content was determined at 205 nm, considering a molar absorptivity constant of 105 L/g.cm (Dence 1992). Soluble lignin content was determined following the protocol described in Dence (Dence 1992). Moisture, extractives, and ash contents were determined according to TAPPI T204 cm-97(Technical Association of Pulp and Paper Industry 1997) and TAPPI T211 (Method and The 1993) protocols, respectively.

Fourier transformed infrared spectroscopy (FTIR)
FTIR analyses were carried out on the fresh pineapple peel, AT, 1B, and 2B to evaluate the samples functional groups. The FTIR spectra were determined using a Perkin Elmer spectrophotometer (Spectrun 100 model), operated with an attenuated total re ection (ATR) and in the transmittance method. Values were measured from 650 to 4000 cm -1 range with 12 scans.

X-ray diffraction (XRD)
The diffraction pattern of the fresh pineapple peel, AT, 1B, and 2Bs were analyzed by X-ray diffraction (XRD, SHIMADZU, Model XDR 6000) with a scanning radiation at 30 kV and 15 mA. All samples dried at 30 °C were scanned from a 2θ of 0° to 60°, with scan speed 0,001-0,002 min -1 . The crystallinity index (CrI) of the samples was calculated using the diffraction intensities of the crystalline structure and that of the amorphous fraction according to Segal's method. (L. Segal et al. 1959) Where: I (002) is the intensity of the crystalline peak at the maximum at 2θ between 22° and 23° for cellulose I, and I (am) is the intensity at the minimum at 2θ between 18° and 19° for cellulose I.

Thermal behavior
The thermal behavior of the PPF, AT, 1B and 2B bers were analyzed on SII Nanotechnology INC equipment (Exstar 6000 model, TG/DTA 6200 series), operating under 100 mL/min constant nitrogen ow, in a temperature range of 30 to 750 ºC and a 10 ºC/min heating rate, and ~ 10 mg of oven-dried samples. For kinetic calculation, three additional heating rates (5, 20, and 40 °C.min -1 ) were included. The kinetic study was carried out using a free Software (Kinetic Calculation v. 1.0 program) and considering Vyazovkin and Flynn-Wall-Ozawa methods (Drozin et al. 2020). The program Kinetic Calculations makes it possible to calculate kinetic substance decomposition parameters for thermogravimetric analysis through the Vyazovkin and the Ozawa-Flynn-Wall methods. More details in the kinetic calculation can be found on (Drozin et al. 2020). Only the activation energy values will be considered in this study for comparison among the samples.

Scanning electron microscopy (SEM)
The morphology of all the bers were characterized by scanning electron microscopy (SEM) using a JEOL JSM 5319 equipment, with tungsten filament operating at 20kV, secondary electron detector, working at a distance of 15 mm and employing a low-vacuum technique. All ber samples were coated with gold before SEM analysis.

Results And Discussion
3.1 Morphology and chemical composition After the rst bleaching step (1B), a yellowish-white color is found, and after the second bleaching step (2B), a completely white color is noted. The variation in the color of PPF with the chemical treatments re ects the changes in chemical composition, with the removal of non-cellulosic materials such as lignin, hemicellulose, extractives, ashes, and oils. The same trend was observed in our previous studies (Cristina et al. 2020;Henrique et al. 2020b;Pereira et al. 2021).
The morphology of the bers were investigated using SEM (Fig. 1a-m). Figure 1(a) shows that fresh pineapple peel had irregular morphologies with different particle sizes, massive wrinkled and depressed blocks (i.e., a kind of akiness) as seen with a greater magni cation (Fig.1b-c), a dense structure because non-cellulosic components such as pectin, lignin, and hemicellulose were intact.Similar morphologies structures have been reported by . Meanwhile, AT resulted in bers with a cleaner and smoother surface structure (Fig 1(d)). After AT, the cellulose bers seem to be more exposed due to the removal of a percentage of amount of hemicellulose and lignin ( Fig.1 e-f), as can be seen in the chemical analysis results as shown in Table 2. Hence, two subsequent treatment with step of bleaching to was necessary to remove these non-cellulosic materials in order to obtain a highly-pure cellulose. Fig. 1(g) indicates that a homogeneous structure remained after 1B, which are shown by various agglomeration of macro brils with elongated walls (Fig. 1(h) and Fig. 1(i)). Finally, in Fig.1(j-m), a brous structure composed of micro bers is easily noted with the ber bundles and elongated rod forms.
The valorisation of pineapple waste into value-added products is in uenced by its chemical composition. Table 1 displays the chemical composition of fresh pineapple peel with a comparison to others studies. The natural bers are composed primarily of carbohydrate polymers (cellulose and hemicellulose), aromatic polymers (mainly lignin), and ashes. However, the composition of cellulose, hemicellulose, lignin, and ashes content of the pineapple peel might vary depending upon the season, variety, maturity period, and location of the cultivation of the pineapple (Banerjee et al. 2018). The fresh pineapple peel in the current study consisted of 24.1% cellulose, 29.3% hemicellulose, 6.3% lignin, and 5.0% ash. Ash 5.05 ±0.10 0.6 9.9 1.5 3.8 After AT, the content cellulose increased to 65.2%, while hemicellulose (19.8%), lignin (3.1%), and ash (2.0%) decreased as expected (Table 2), indicating the effectiveness of sodium hydroxide treatment. For further removal of the non-cellulosic materials still remaining in the ber after AT, these bers were subjected to two alkaline bleaching steps (Table 2).

FTIR Analysis
FTIR was used to determine the main functional groups of pineapple peel fresh and the changes caused by the chemical treatments (Fig. 2). In the spectra, two common absorbance regions can be observed, one from 750 to 1750 cm -1 and the other from 2250 to 3850 cm -1 . A similar pattern is observed for all samples. For example, no appearance/disappearance of bands is noted, but an increase in the polysaccharide group intensity (~1300-900 cm −1 ) can be seen. The main bands appeared at 3330, 2900, 1730, 1640, 1512, 1200-1300, and 894 cm −1 .
The 3300 cm −1 and 2900 cm -1 bands correspond to the hydroxyl groups (inter, intra and free OH) (Dai and Fan 2011;Ornaghi et al. 2014) and to stretching vibration of methyl and methylene. The band at 1730 cm −1 is assigned to C O in uronic acid. A similar pattern is observed among the samples, i.e., no appearance/disappearance of bands is noted, but an increase in the polysaccharide group intensity (~1300-900 cm -1 ) can be seen. Also, a slight increase in the OH region intensity is observed for all treatments compared to the spectrum of the PPF. The PPF spectrum assigns some characteristic bands 1730, 1640, and 1512 cm −1 , also referred to hemicellulose and lignin. The 1730 cm −1 band peak is assigned to carbonyl stretching in unconjugated ketones and conjugated carbonyl groups (de Menezes Nogueira et al. 2019). It can be attributed either to the acetyl and uronic ester groups of hemicellulose or to the ester linkage of the carboxylic groups of ferulic and p-coumeric acids of lignin and/or hemicelluloses (Johari et al. 2016). The bands at 1512 cm −1 and 1640 cm −1 represent the aromatic C=C of lignin's aromatic rings (Hoareau et al. 2004). The band at 1253 cm −1 corresponds to C-O-C (aryl-alkyl ether (Romanzini et al. 2013). This peak is generally observed when the guayacil ring, a subunit of the lignin macromolecule, is present . All samples showed some characteristic signature bands of cellulose, including 3300, 2902, 1028, and 894 cm −1 , which corresponds to O-H stretching, C-H stretching, C-H deformation, C-O-C pyranose ring stretching vibration, and β-glycosidic linkages, respectively. As aforementioned, all these bands seem to slightly increase after the treatments (and maintaining a similar pattern) in comparison to PPF. The bands at 894 cm −1 for all samples are characteristic of β-glycosidic linkages between sugar molecules (Banerjee et al. 2019).

XRD Analysis
The diffractograms of each of the analysed samples (AT, 1B, and 2B) are present in Fig.3. The peaks at 2θ=14.8° and 2θ=16.3° corresponds to two overlapped weaker diffraction assigned of the (1-10) and (110) lattice planes (Besbes et al. 2011), a sharp peak at 2θ=22.5°, which is assigned to the (200) lattice (French 2014a), and small and broad peak at 2θ = 34.5° represents the contribution of (040) plane (French 2014b;Tonoli et al. 2021). Only one peak (2θ = 22.5) is observed in the re ective pattern of the fresh pineapple peel. This could be due to the presence of amorphous materials covering the crystalline portion of the ber. After alkaline treatment and bleaching of the fresh pineapple, to obtain AT, 1B and 2B two new crystalline peaks emerged in the diffractogram of at 2θ = 15.4° and 34.5°. These peaks become more de ned upon chemical treatments as expected. 2B, therefore, show 3 peaks at 15.4°, 22.5° and 34.5° similar to what has been reported for the (110), (200) and (004) crystallographic planes of cellulose I polymorph thus indicating that 2B has typical cellulose I structure (French and Santiago Cintrón 2013).
XRD analysis was conducted to provide the effect of the chemical treatments on the Segal empirical's crystallinity index (CrI.) (L. Segal et al. 1959). The CrI. was approximately 24.05, 53.74, 79.00, and 80.91% for PPF, AT, 1B, and 2B bers, respectively. The crystallinity increased signi cantly from 24.05% to 80.91.8%. The higher crystallinity index of PPF compared to 1B (24.05 vs. 79%) could be well understood as a result of the removal of amorphous non-cellulosic compounds, induced by the puri cation treatment (alkali and bleaching treatments) performed to purify cellulose, which is due to the effective removal of amorphous components such as lignin, hemicellulose, waxes, and pectin with the treatments.

Thermogravimetric (TG) and thermal degradation kinetic
The TG and DTG curves of PPF before and after each chemical treatment (AT, 1B, and 2B) are shown in Fig (4a and 4b), with heating rate 10°Cmin -1 . The degradation temperatures for all samples can be observed in Table 3.
In general, the thermal decomposition of PPF can be divided into four main stages. The rst stage corresponds to the temperature range of 30-140°C, and the mass loss is mainly attributed to moisture evaporation and low molecular mass components degradation . This stage is observed for all samples, independently of the treatment. The other stages of mass loss are attributed to hemicellulose, cellulose, and lignin components. For the treated bers, only one peak referred to as cellulose is observed, showing that the remained non-cellulosic components have a small in uence on the degradation stages. All treated curves showed similar behavior. The second stage corresponds to the decomposition of hemicellulose and part of cellulose (cleavage of cellulose glycosidic bonds) (Hu et al. 2020;Marques et al. 2020). Finally, lignin degrades in a wide temperature range (100-900 °C) (Melikoğlu et al. 2019;Ornaghi et al. 2020).
For 1B and 2B bers, the more pronounced mass loss was observed between 300 and 400 °C. The major cellulosic chain degradation is due to several steps, such as depolymerization, dehydration, and decomposition of the glycosidic units (Gabriel et al. 2020). Also, the thermal stability of bers bleached increases after each chemical treatment (Table 3). The activation energies were calculated using free available Software (Kinetic Calculation v. 1.0 program) (Drozin et al. 2020) and are presented in Figure 5. The conversion range (α) from 0.2 to 0.8 was used according to the recommendation in Drozin 2020 (Drozin et al. 2020) due to the inaccuracy of the process's mathematical description. Vyazovkin and FWO methods were used to estimate the Ea. In general, all treatments gave a more linear activation energy distribution than the PPF, which increases with conversion degree. The values obtained are presented in Table 4. A relative error higher than 10% indicates a multi-step mechanism degradation, which only occurs for PPF. In general, the linearity of distribution of activation energy can be attributed to the predominant presence of cellulose (indicating a single-step mechanism degradation), which increased by treatment due to removal of more impurities and non-cellulosic components, i.e., in this case, higher Ea was proportional to the crystallinity index with the exception of the PPF.
It is important to mention (and can be counterintuitive) that the higher crystallinity for the treated samples (as observed in XRD, morphology, and chemical composition) does not necessarily means higher activation energies due to the complex pyrolysis behavior of this type of materials. Hence, the higher crystallinity for lignocellulosic materials (mainly in our case) is indicative of a more homogeneous E a distribution and not higher values per se.

Conclusions
In this study, cellulose bers were successfully extracted from pineapple peels characterized in terms of physical, chemical, thermal, and morphological properties. The high cellulose content and low hemicellulose after the alkaline treatment sequence lead to cellulose, which has great potential applications in the automotive industry as reinforcement polymeric, and produce cellulose nanocrystal to use as food packaging material. Three-stage extraction was bene cial to maintain the morphology of cellulose bers, which produced cellulose I with high purity and crystallinity (80.91 %). The alkaline treatment showed higher e ciency in removing the lignocellulosic compound, as it was con rmed via FTIR and compositional analyses. The kinetic parameters of the degradation of PPF, AT, 1B, and 2B were accurately determined from a series of TGA experiments at four heating rates. The Ea value of the degradation of the PPF calculated using Vyazovkin and Flynn-Wall-Ozawa methods based on TGA and DTGA curves are higher than those of AT, 1B, and 2B materials. The activation energy required for initiating the thermal degradation after alkaline treatment was less than the value calculated to 2B due to quantities de hemicellulose and lignin in the structure. The calculated average apparent activation energy value through Vyazovkin and FWO methods were 175.52 and 195.58 kJ mol −1 for bleached cellulose (2B), while to AT were 168.17 and 184.87 kJ mol −1 , respectively. From the above conclusion, it can say it's possible to conversion of fruit residue into a high-value product and generated new information about the speci c treatment conditions, bleaching condition for obtaining high-quality cellulose, and environmentally friendly cellulose extraction methods for applications in diverse areas.